Hyper Scale-Out Platform. Using the Management Console

Transcription

1 Hyper Scale-Out Platform MK-94HSP November 2016

2 2016 Hitachi, LTD. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for commercial purposes without the express written permission of Hitachi, Ltd., or Hitachi Data Systems Corporation (collectively, Hitachi ). Licensee may make copies of the Materials provided that any such copy is: (i) created as an essential step in utilization of the Software as licensed and is used in no other manner; or (ii) used for archival purposes. Licensee may not make any other copies of the Materials. "Materials" mean text, data, photographs, graphics, audio, video and documents. Hitachi reserves the right to make changes to this Material at any time without notice and assumes no responsibility for its use. The Materials contain the most current information available at the time of publication. Some of the features described in the Materials might not be currently available. Refer to the most recent product announcement for information about feature and product availability, or contact Hitachi Data Systems Corporation at Notice: Hitachi products and services can be ordered only under the terms and conditions of the applicable Hitachi agreements. The use of Hitachi products is governed by the terms of your agreements with Hitachi Data Systems Corporation. By using this software, you agree that you are responsible for: 1) Acquiring the relevant consents as may be required under local privacy laws or otherwise from authorized employees and other individuals to access relevant data; and 2) Verifying that data continues to be held, retrieved, deleted, or otherwise processed in accordance with relevant laws. Notice on Export Controls. The technical data and technology inherent in this Document may be subject to U.S. export control laws, including the U.S. Export Administration Act and its associated regulations, and may be subject to export or import regulations in other countries. Reader agrees to comply strictly with all such regulations and acknowledges that Reader has the responsibility to obtain licenses to export, re-export, or import the Document and any Compliant Products. Hitachi is a registered trademark of Hitachi, Ltd., in the United States and other countries. AIX, AS/400e, DB2, Domino, DS6000, DS8000, Enterprise Storage Server, eserver, FICON, FlashCopy, IBM, Lotus, MVS, OS/390, PowerPC, RS6000, S/390, System z9, System z10, Tivoli, z/os, z9, z10, z13, z/vm, and z/vse are registered trademarks or trademarks of International Business Machines Corporation. Active Directory, ActiveX, Bing, Excel, Hyper-V, Internet Explorer, the Internet Explorer logo, Microsoft, the Microsoft Corporate Logo, MS-DOS, Outlook, PowerPoint, SharePoint, Silverlight, SmartScreen, SQL Server, Visual Basic, Visual C++, Visual Studio, Windows, the Windows logo, Windows Azure, Windows PowerShell, Windows Server, the Windows start button, and Windows Vista are registered trademarks or trademarks of Microsoft Corporation. Microsoft product screen shots are reprinted with permission from Microsoft Corporation. All other trademarks, service marks, and company names in this document or web site are properties of their respective owners.

3 Contents Preface Intended audience Product version Syntax notation Release notes Accessing product documentation Getting help Comments vii vii vii vii viii ix ix ix Chapter 1: Performing initial cluster configuration 1 Chapter 2: 9 Layout and Navigation 9 Resource selection and associated tasks 9 Context-sensitive online help 10 Obtaining HSP version information 10 Chapter 3: Monitoring event logs 11 Chapter 4: Monitoring jobs 13 Chapter 5: Monitoring performance 15 Chapter 6: Monitoring alerts 17 Chapter 7: Managing the properties of a cluster 21 Viewing and editing the cluster properties 22 Viewing and editing the file storage properties 24 Contents iii

4 Viewing and editing the network properties 27 Viewing and editing the SMTP properties 28 Managing IP addresses 31 Viewing the properties of an IP address 32 Adding an IP address 34 Editing the properties of an IP address 37 Deleting an IP address 39 Viewing and editing options 40 Chapter 8: Managing virtual machine resources 41 Managing VM templates 43 Viewing the properties of a VM template 45 Adding a VM template 47 Editing the properties of a VM template 49 Cloning a VM template 50 Deleting a VM template 52 Managing VM instance groups 53 Viewing the properties of a VM instance group 54 Adding a VM instance group 55 Adding VM instances to a VM instance group 58 Viewing the list of VM instances associated with a VM instance group 60 Deleting VM instances from a VM instance group 61 Editing the properties of a VM instance group 62 Powering on the VM instances in a VM instance group 63 Shutting down the VM instances in a VM instance group 64 Deleting a VM instance group 65 Managing VM instances 66 Viewing the properties of a VM instance 67 Adding VM instances 70 Powering on a VM instance 73 Working with a VM instance in sandbox mode 74 Editing the properties of a VM instance 76 Migrating a VM instance 77 Managing VM instance snapshots 78 Listing the VM snapshots for a VM instance 79 Adding a VM snapshot 80 Editing a VM snapshot 81 Reverting a VM snapshot 82 Deleting a VM snapshot 83 iv Contents

5 Attaching a disk to a VM instance 84 Detaching a disk from a VM instance 85 Rebooting a VM instance 86 Shutting down a VM instance 87 Deleting a VM instance 88 Managing VM sizes 89 Viewing the properties of a VM size 90 Adding a VM size 92 Editing the properties of a VM size 93 Deleting a VM size 94 Managing VM volumes 95 Viewing the properties of a VM volume 96 Adding VM volumes 98 Editing the properties of a VM volume 99 Deleting a VM volume 100 Chapter 9: Managing cluster resources 101 Managing tenants 102 Viewing the properties of a tenant 104 Adding a tenant 106 Editing the properties of a tenant 107 Deleting a tenant 108 Managing file systems 109 Viewing the properties of a file system 111 Adding a file system 116 Enabling or disabling a file system 121 Editing the properties of a file system 122 Checking and repairing a file system 126 Deleting a file system 127 Managing shares 128 Viewing the properties of a share 129 Adding a share 130 Editing the properties of a share 132 Managing access rules 134 Viewing the list of access rules for a share 135 Adding an access rule 136 Editing an access rule 140 Deleting an access rule 144 Deleting a share 145 Contents v

6 Managing nodes 146 Viewing the properties of a node 148 Editing the properties of a node 151 Identifying a node 153 Changing node rack assignment 154 Putting a node in maintenance mode 155 Rebooting a node 156 Shutting down a node 157 Deleting a node 158 Managing racks 159 Viewing the properties of a rack 160 Adding a rack 161 Editing the properties of a rack 162 Deleting a rack 163 Managing disks 164 Viewing the properties of a disk 166 Editing the properties of a disk 168 Identifying disks 170 Deleting a disk 171 Managing user accounts 172 Viewing the properties of a user account 173 Adding a user account 175 Editing the properties of a user account 178 Deleting a user account 181 Index 183 vi Contents

7 Preface Provided for your convenience, this document is a printable PDF version of the online help that is embedded in the Management Console GUI. Intended audience This document is intended for anyone new to HSP and needs an high-level overview of the product. Product version This document applies to Hyper Scale-Out Platform release or later. Syntax notation The table below describes the conventions used for the syntax of commands, expressions, URLs, and object names in this document. boldface italic Convention <monospace italic> in angle brackets Indicates text in a user interface window or dialog box, such as menus, menu options, buttons, and labels. For example: On the Add Share dialog box, click OK. Indicates important new terms. Indicates a variable, which is a placeholder for site- or installationspecific details that you need to provide. For example: Copy <source-file> <target-file> Preface vii

8 Release notes (Continued) Convention monospace Indicates: The name of a directory, folder, or file. For example: The nodes.xml file Text that is displayed on the screen. For example: # cluster_initialize monospace bold <monospace italic> in angle brackets Indicates text you enter. For example: $ cluster_initialize Indicates a variable, which is a placeholder for site- or installationspecific details that you need to provide. For example: Copy <source-file> <target-file> [ ] square brackets Indicates optional values. For example: [ a b ] indicates that you can choose a, b, or nothing. { } curly braces Indicates required or expected values. For example: { a b } indicates that you must choose either a or b. vertical bar Indicates that you have a choice between two or more options or arguments. For example: [ a b ] indicates that you can choose a, b, or nothing. { a b } indicates that you must choose either a or b. [ ] Indicates optional values. For example: [ a b ] indicates that you can choose a, b, or nothing. ellipsis Indicates text that was removed to shorten the example output. Release notes Read the release notes before installing and using this product. They may contain requirements or restrictions that are not fully described in this document or updates or corrections to this document. Release notes are available on Hitachi Data Systems Support Connect: viii Preface

9 Accessing product documentation Accessing product documentation Product documentation is available on Hitachi Data Systems Support Connect: Check this site for the most current documentation, including important updates that may have been made after the release of the product. Getting help Hitachi Data Systems Support Portal is the destination for technical support of products and solutions sold by Hitachi Data Systems. To contact technical support, log on to Hitachi Data Systems Support Connect for contact information: Hitachi Data Systems Community is a global online community for HDS customers, partners, independent software vendors, employees, and prospects. It is the destination to get answers, discover insights, and make connections. Join the conversation today! Go to community.hds.com, register, and complete your profile. Comments Please send us your comments on this Help: hsp.documentation.comments@hds.com Include the Help title and part number, and refer to specific topics and paragraphs whenever possible. All comments become the property of Hitachi Data Systems. Thank you! Preface ix

10 x Preface

11 1 Performing initial cluster configuration When you access the HSP Management Console for the first time, the software detects that you need to configure the cluster and displays the Initial Cluster Configuration page. You will not be able to perform any other Management Console management tasks until you have successfully configured the cluster. A companion document, the Configuration Checklist, can aid you in collecting and organizing the information required to configure the HSP cluster. To configure your cluster 1. Launch the Google Chrome Web browser. 2. Enter the virtual IP address of the cluster in the address bar of the browser. For example: 3. Log into the Management Console using the User Name of admin and a Password of admin. The HSP software automatically creates the administrative account called admin. You will be prompted for a new password for this account during initial configuration. The HSP software detects that no configuration has yet been performed for the cluster and opens the Cluster Properties page. Chapter 1: Performing initial cluster configuration 1

12 4. Accept the default values provided or specify values for the following cluster configuration properties: Property Name Name of the cluster. The cluster is configured at the factory with a default cell name of "MyCluster." You can, however, change the name of your cluster. ID Virtual IP Password for Admin User Cluster's universally unique identifier (UUID). Virtual IP address through which clients access file systems and administrators access the management interfaces within a cluster. There is only one virtual IP address associated with a cluster. The installer automatically creates an administrative account called admin with a password of admin. You are required to provide a new password for this account during configuration. The password must be a minimum of 5 and a maximum of 255 characters and can contain any UTF-8 characters except ; (semicolon). You cannot specify the initial default password of admin. Confirm Password for Admin User Version Tags Must match what you entered in Password for Admin User. Version of HSP software installed on a nodes in the cluster. The first three period-separated numbers (for example, 1.2.2) are the version number, and the digits after the last period following the version number are the build number. of the cluster. User-defined tags that can be used to organize and/or filter the response results. 5. Click Next. The File Storage Properties page is displayed. 2 Chapter 1: Performing initial cluster configuration

13 6. Accept the default values provided or specify values for the following file storage configuration properties: Property Storage Space HWM Storage space usage high water mark percentage defined for the cluster. When the remaining cluster storage space usage reaches the high water mark percentage, an alert is raised until the storage space usage falls below the low water mark percentage. By default, the storage space high water mark is set to 80. Storage Space LWM Storage space low water mark percentage defined for the cluster. By default, the storage space low water mark is set to 70. Existing storage space usage alerts are automatically cleared once the low water mark percentage is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. File System Space HWM High water mark percentage set by default for a new file system added to a cluster. You can override this percentage at the file system level. When the remaining allocated file system space reaches the high water mark, an alert is raised until the file system space consumption falls below the low water mark percentage. By default, file system space high water mark is set to 90. File System Space LWM Low water mark percentage set by default for a new file system added to a cluster. Existing file system space alerts are automatically cleared once the low water mark setting is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. By default, the file system space low water mark is set to 85. Chapter 1: Performing initial cluster configuration 3

14 (Continued) Property File System Auto Access Whether or not the required share and access rule will be automatically created when a file system is added. By default, File System Auto Access is set to Yes. When you add a file system, the HSP software adds a share with the same name as the file system and adds an access rule called "DefaultAccessRule" that provides read/write access to the share for all client hosts. You can add more restrictive access rules to the share, as well as edit or delete the automatically created "DefaultAccessRule." Hide Shares Whether or not share visibility is restricted only to clients that have permission to access the share. By default, Hide Shares is set to No. 7. Click Next. The Network Properties page is displayed. 8. Accept the default values provided or specify values for the following network configuration properties: Property Gateway IP address of your network gateway. This is used by HSP to access external services (for example, NTP for time synchronization and SMTP for notification delivery). Domain Name DNS Server NTP Server Your company s valid domain name, for example: YourCompanyName.com IP address of your DNS server. IP address or fully qualified host name of your NTP server. Network Time Protocol (NTP) is used to synchronize time on the nodes within a cluster. External Floating IP Address Pool The pool of IP addresses that are assigned to the nodes by the HSP software. We call these floating IP addresses because they can migrate or float from one node to another to provide continuous service when individual nodes go down and up. As such, you do not explicitly assign an external address to each node. 4 Chapter 1: Performing initial cluster configuration

15 (Continued) Property Important! You should maintain at least as many IP addresses as there are nodes in the cluster twice the number of IP addresses per node is recommended. Also, you cannot assign IP addresses to an HSP cluster that are already in use within your network. IP addresses are specified in IPv4 CIDR notation. You can specify the IP addresses as: A range of IP addresses, for example: / /21 A range of IP addresses using a special notation for contiguous addresses, for example: /21+10 which would generate IP addresses /21 through /21. A comma-separated list of IP addresses, for example: /21, /21, /21 The HSP software automatically generates a name for each IP address, which is a comma-separated version of the IP address (for example /21 becomes ). Optionally, you can also specify a prefix that will be added to the name. For example, if you specify a prefix of floating, then /21 becomes floating Virtual Machine IP Address Name Prefix The HSP software automatically generates a name for each IP address, which is a comma-separated version of the IP address (for example /21 becomes ). Optionally, you can also specify a prefix that will be added to the name. For example, if you specify a prefix of hadoop, then /21 becomes hadoop Chapter 1: Performing initial cluster configuration 5

16 (Continued) Property Virtual Machine IP Address Pool The pool of IP addresses that are assigned by the HSP software to VM instances when a VM template is deployed. Notes: You need at least as many IP addresses as the number of instances you want to deploy. Also, you cannot assign IP addresses to an HSP cluster that are already in use within your network. IP addresses are specified in IPv4 CIDR notation. You can specify IP addresses as: A range of IP addresses, for example: / /21 A range of IP addresses using a special notation for contiguous addresses, for example: /21+10 which is the same as specifying the range above. A comma-separated list of IP addresses, for example: /21, /21, /21 9. Click Next. The SMTP Properties page is displayed. You can configure SMTP now, or you can finish cluster configuration and configure SMTP at a later time. Go to the next step to configure SMTP properties or click Next. 10. Accept the default values provided or specify values for the following SMTP configuration properties: Property Name Name of the SMTP configuration. By default, the name of the existing SMTP configuration is "SMTPConfig," but you can change the name. SMTP configuration names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). 6 Chapter 1: Performing initial cluster configuration

17 (Continued) Property Enabled Whether or not the SMTP configuration is enabled. By default, the initial default SMTP configuration is disabled because the SMTP configuration requires editing for your environment. When you enable the SMTP configuration, you must provide the required SMTP server and address properties. SMTP Server (required) SMTP Server Port Display Name IP address or fully qualified host name of the SMTP server. For example: "smtp.mailserver.com" If specified as an IP address, use IPv4 notation, for example: The outgoing SMTP server port. The default port for a non-ssl configuration is 25. The default port for an SSL configuration is 465. The from display name associated with sent alert messages. For example: "admin." Display names can be from 1 to 255 characters and can contain any UTF-8 characters except ; (semicolon). Address (required) Security Enabled The from address associated with sent alert messages. For example: "WestCoastCluster@hds.com" Unicode characters are supported. Whether or not the client should use Secure Socket Layer (SSL) based security when sending . By default, Security Enabled is set to No. If you use SSL, you must provide a user name and password that will be passed to the service provider for authentication. User Name User name that is passed to the service provider for authentication. Required if Security Enabled is set to Yes. Password Password that is passed to the service provider for authentication. Required if Security Enabled is set to Yes. Chapter 1: Performing initial cluster configuration 7

18 (Continued) Property of the SMTP configuration. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 11. Click Next. The Confirm Cluster Configuration page is displayed. 12. Review the cluster configuration values you specified during configuration and edit the properties as required. 13. Click Finish to configure the cluster or return to previous pages to make configuration changes. 14. Cluster configuration takes a few minutes to complete. You will then need to log into the Management Console using the administrative user account admin and the password you supplied during configuration. 8 Chapter 1: Performing initial cluster configuration

19 2 After successfully logging into the HSP Management Console, you are presented with the Dashboard page. Take a few minutes to familiarize yourself with the basic layout of the interface and the location and meaning of important navigation controls and visual cues. Layout and Navigation Left side navigation provides access to the manageable HSP resources. Each entity that you can manage independently in the Management Console is referred to as a resource. Top navigation provides access to the action-based aspects of managing the HSP cluster, including monitoring cluster events, jobs, performance, and alerts. Across the top of the dashboard window are tiles that display alerts for Hardware and File System. If the cluster is healthy, a check mark is displayed in the center of the alert tiles. If the HSP software detects problems, a number displayed in the center of the alert tiles summarizing the number of problems detected. Click an alert tile to display problem details. Circular information gauges provide summaries of file system space usage and virtual machine resource consumption. Resource selection and associated tasks The Management Console offers two different ways to display and manage resources: Tile view Chapter 2: 9

20 Context-sensitive online help Click the name of the resource for details. List view Click the ellipses (...) under Details in any row for details. Icons above the tiles or list Management tasks represent the resource-specific management tasks you can perform on a resource. Breadcrumb drop-down menus provide navigation when you are in a resource page. Context-sensitive online help Click to obtain to access contextual page-level online help. Obtaining HSP version information Click to obtain the software version and build information, should you need to contact Hitachi Data Systems Support. The first three periodseparated numbers (for example, 1.2.0) are the version number, and the digits after the last period following the version number are the build number. 10 Chapter 2:

21 3 Monitoring event logs Event logs include all internal system generated messages (informational and notice activities, as well as alerts) and administrator-initiated activities that occur within the cluster. The HSP software maintains about 10,000 event logs. Older event logs are archived in the system_archive file system. Event logs that are associated with alert conditions are not archived until the system clears those alert conditions. Important: To access archived alert conditions, you must create hostbased access rules that are appropriate for your environment for the system_archive. By default, system_archive is not accessible. See "Adding an access rule" on page 136 for details. 1. Click Events in the top navigation. 2. Select an event to view the details: Property Name ID Severity Resource Name Resource Type Timestamp Resource ID Name of the event log. Event log's universally unique identifier (UUID). Severity of the event info, debug, critical, error, or warning. Name of the resource to which the event log is associated. Type of the resource to which the event log is associated. Date and time that the event occurred. ID of the resource to which the event log is associated. Chapter 3: Monitoring event logs 11

22 (Continued) Tags Property of the event log. User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 12 Chapter 3: Monitoring event logs

23 4 Monitoring jobs Because of the multi-threaded and individual task queueing nature of the HSP technology, the Management Console lets you track the status of each individual job associated with an asynchronous management operation. Adding, editing, and deleting a resource are all asynchronous operations, which lets you continue to perform management tasks without waiting until these operations complete. Asynchronous operations are queued up as "jobs" and are run in the background. You can track the status of these jobs from the Jobs page. A red revolving circle appears in the top navigation bar next to Jobs when a job is submitted. A red triangle appears in the top navigation bar next to Jobs if there are job failures that you have not yet viewed on the Jobs page. The number inside the triangle represents the number of jobs that have failed since you last viewed the Jobs page. 1. Click Jobs in the top navigation. 2. Select a job to view the details: Property Resource Name ID Resource Type Creation Timestamp Name of the resource upon which the POST or DELETE operation is being performed. Job's universally unique identifier (UUID). Type of resource upon which the POST or DELETE operation is being performed, for example a POST on a cluster. Job creation timestamp. Chapter 4: Monitoring jobs 13

24 (Continued) Property Completion Status Completion Details Completion Substatus Completion status of the asynchronous operation: PROCESSING COMPLETE ERROR Additional information about the completion status of the asynchronous operation. Additional information about the completion status of the asynchronous operation: For the completion-substatus of COMPLETE: OK For the completion-substatus of PROCESSING: NOT_QUEUED QUEUED RUNNING For the completion-substatus of ERROR: RETRY FAIL INVALID UNKNOWN Resource Action Percent Complete Resource ID Target Node Name Target Node ID Spawned Jobs Spawned Jobs List The action being performed: ADD EDIT DELETE EXECUTE Completion percentage of the asynchronous operation. Resource identifier upon which the POST or DELETE operation is being performed. ID of the node targeted to run the job. ID of the node targeted to run the job. Whether or not additional jobs were spawned to accomplish the job. List of URLs to additional jobs that were spawned to accomplish the task requested in this job. 14 Chapter 4: Monitoring jobs

25 5 Monitoring performance You can create performance graphs representing the data that the HSP software collects and maintains for: Load Memory usage CPU usage Network traffic Note: Keep in mind that historical data is only available from the time your cluster was created and brought online. 1. Click Performance in the top navigation. 2. Select either the cluster or a particular node for which to graph performance from the drop-down. 3. Select the Metric you want to graph from the drop-down. 4. Select the Time Period you want to graph from the drop-down. 5. Click Generate Graph to create the graph. Change the options and click Generate Graph to create a new performance graph. Chapter 5: Monitoring performance 15

26 16 Chapter 5: Monitoring performance

27 6 Monitoring alerts Alert conditions are issues or negative circumstances that may require customer support or administrative attention and/or corrective action. Alerts are automatically removed by the HSP software once the underlying issue that caused the alert condition is resolved. The following summarizes the alert conditions that are raised for the various HSP resources: Cluster Warning severity alert conditions are raised when the administrator-defined high and low watermark thresholds for the storage space of the cluster are out of the specified range. Node Warning severity alert conditions are raised when the run state property of a node is DOWN, the node has been put in maintenance mode, or memory consumption may cause the node to be rebooted. Error severity alert condition is raised when the run state property of a node is ERROR. Disk Error severity alert condition is raised when the run state property of a disk is ERROR. IP address Warning severity alert condition is raised when the run state property of an IP address is DOWN. File system Warning severity alert conditions are raised when the administrator-defined high and low watermark thresholds for file system space are out of the specified range. Warning severity alert condition is raised when the run state property of a file system is IMPAIRED. Error severity alert condition is raised when the run state property of a file system is ERROR. Alert conditions are also raised when a node s power supply, fan speed, and CPU temperature sensors detect the following: Chapter 6: Monitoring alerts 17

28 Power supplies Warning severity alert condition is raised when a power supply cord on a node has been unplugged. Error severity alert condition is raised when a power supply has failed. Sensors Warning severity alert conditions are raised when the thresholds that have been set for them have been exceeded. To send notifications to HSP administrators about alert conditions, you need: A valid SMTP configuration. Review and edit the SMTP configuration. See "Viewing and editing the SMTP properties" on page 28 for more information. To configure alert condition notifications within a HSP user account. See "Editing the properties of a user account" on page 178 for more information about subscribing to critical, error, or warning alert conditions. Cleared alerts are archived in the system_archive file system share. The most recently cleared 150 (approximate) alerts are maintained in this archive. Important: To access archived alert conditions, you must create hostbased access rules that are appropriate for your environment for the system_archive. By default, system_archive is not accessible. See "Adding an access rule" on page 136 for details. Across the top of the dashboard window are tiles that display alerts for Hardware and File System. If the cluster is healthy, a check mark is displayed in the center of the alert tiles. If the HSP software detects problems, a number displayed in the center of the alert tiles summarizing the number of problems detected. Click an alert tile to display problem details. 1. Click either the Hardware or File System tile. 2. Select an alert to view the details: Property Name Name of the alert condition. 18 Chapter 6: Monitoring alerts

29 (Continued) Property ID Severity Resource Name Resource Type Time Raised Resolution Resource ID Tags Alert condition's universally unique identifier (UUID). Severity of the alert condition warning, error or critical. Name of the resource to which the alert condition is associated. Type of the resource to which the alert condition is associated. Date and time that the alert condition was raised. describing the administrative action that should be considered to correct or clear the alert condition. ID of the resource to which the alert condition is associated. of the alert condition. User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 6: Monitoring alerts 19

30 20 Chapter 6: Monitoring alerts

31 7 Managing the properties of a cluster Clusters are organizational units that allow for connectivity and data segmentation within HSP. A cluster consists of a minimum of five nodes that are physical machines that provide collective storage for the cluster. You can edit the cluster properties at any time, however, exercise caution when making modifications to the core properties of the cluster. Many of these properties are set to the recommended defaults or were assigned based on your site-specific infrastructure when the cluster was initially configured. Important: Changes that you make to these properties could impact cluster health, as well as internal and external cluster communications. You can perform the following cluster management tasks: Viewing and editing the cluster properties Viewing and editing the file storage properties Viewing and editing the network properties Viewing and editing the SMTP properties Chapter 7: Managing the properties of a cluster 21

32 Viewing and editing the cluster properties Viewing and editing the cluster properties 1. In the top right, click. 2. Select Cluster Properties. 3. Review the configuration values and click to edit the properties as required: Property Name Name of the cluster. The cluster is configured at the factory with a default cell name of "MyCluster." You can, however, change the name of your cluster. If you change the name of the cluster, add the new name and its corresponding virtual IP address to your local DNS server or to the /etc/hosts files on all of the hosts that need to access the cluster. Cluster names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Virtual IP Cluster's universally unique identifier (UUID). Virtual IP address through which clients access file systems and administrators access the management interfaces within a cluster. There is only one virtual IP address associated with a cluster. Valid IP address specified in IPv4 CIDR notation, for example: /24. Version Version of HSP software installed on a nodes in the cluster. The first three period-separated numbers (for example, 1.2.2) are the version number, and the digits after the last period following the version number are the build number. of the cluster. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 22 Chapter 7: Managing the properties of a cluster

33 Viewing and editing the cluster properties Text input fields are provided for those properties that you can edit. 4. Click OK to save your changes, or Cancel to return to Cluster Properties without making changes. Chapter 7: Managing the properties of a cluster 23

34 Viewing and editing the file storage properties Viewing and editing the file storage properties 1. In the top right, click. 2. Select File Storage Properties. 3. Review the configuration values and click to edit the properties as required: Property Storage Space HWM Storage space usage high water mark percentage defined for the cluster. When the remaining cluster storage space usage reaches the high water mark percentage, an alert is raised until the storage space usage falls below the low water mark percentage. By default, the storage space high water mark is set to 80. Specify a value from 1 to 98 that is greater than the storage space low water mark. Evaluate and adjust the high water mark default to ensure that you are alerted when your cluster is in danger of having insufficient space to maintain data protection policies if nodes were to fail. To adjust the storage space high water mark, decide on the number of node failures to tolerate as a percentage of the total cluster size, then use this equation: HWM = 100% * (1 - (node-failures-to-tolerate total-#-nodes) Using the default as an example, if node-failures-to-tolerate is 1 node failure in a 5 node cluster, then 1 5 would be 20%. In this case the high water mark would be 100% * (1 -.20) or 80%. Here are some other examples: 100% * (1 - (3 10)) = 70% 100% * (1 - (7 100)) = 93% 24 Chapter 7: Managing the properties of a cluster

35 Viewing and editing the file storage properties (Continued) Property Storage Space LWM Storage space low water mark percentage defined for the cluster. By default, the storage space low water mark is set to 70. Existing storage space usage alerts are automatically cleared once the low water mark percentage is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. Specify a value from 1 to 98. File System Space HWM High water mark percentage set by default for a new file system added to a cluster. You can override this percentage at the file system level. When the remaining allocated file system space reaches the high water mark, an alert is raised until the file system space consumption falls below the low water mark percentage. By default, file system space high water mark is set to 90. Specify a value from 1 to 100 that is greater than the file system space low water mark. File System Space LWM Low water mark percentage set by default for a new file system added to a cluster. Existing file system space alerts are automatically cleared once the low water mark setting is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. By default, the file system space low water mark is set to 85. Specify a value from 1 to 100. Chapter 7: Managing the properties of a cluster 25

36 Viewing and editing the file storage properties (Continued) Property File System Auto Access Whether or not the required share and access rule will be automatically created when a file system is added. By default, File System Auto Access is set to Yes. When you add a file system, the HSP software adds a share with the same name as the file system and adds an access rule called "DefaultAccessRule" that provides read/write access to the share for all client hosts. You can add more restrictive access rules to the share, as well as edit or delete the automatically created "DefaultAccessRule." If you set File System Auto Access to No, you will need to manually add a share and at least one access rule to share out a file system. Alternatively, you can set the Auto Access property at file system creation time to override this cluster setting and allow the automatic creation of the share and access rule (). Hide Shares Whether or not share visibility is restricted only to clients that have permission to access the share. By default, Hide Shares is set to No. Specify Yes or No. You can restrict share visibility in multitenant environments by setting Hide Shares to Yes. Clients will then only see the shares for which they have access permissions. Text input fields are provided for those properties that you can edit. 4. Click OK to save your changes, or Cancel to return to File Storage Properties without making changes. 26 Chapter 7: Managing the properties of a cluster

37 Viewing and editing the network properties Viewing and editing the network properties 1. In the top right, click. 2. Select Network Properties. 3. Review the configuration values and click to edit the properties as required: Property Gateway IP address of your network gateway. This is used by HSP to access external services (for example, NTP for time synchronization and SMTP for notification delivery). Valid IP address specified in IPv4 notation, for example: Domain Name DNS Server Your company s valid domain name, for example: YourCompanyName.com IP address of your DNS server. Valid IP address specified in IPv4 notation, for example: NTP Server IP address or fully qualified host name of your NTP server. Network Time Protocol (NTP) is used to synchronize time on the nodes within a cluster. If specified as an IP address, use IPv4 notation, for example: Text input fields are provided for those properties that you can edit. 4. Click OK to save your changes, or Cancel to return to the Network Properties page without making changes. Chapter 7: Managing the properties of a cluster 27

38 Viewing and editing the SMTP properties Viewing and editing the SMTP properties A valid SMTP configuration is required for an HSP user to receive notifications of alert conditions. A valid SMTP configuration is required for an HSP user to receive notifications of alert conditions. The HSP software comes with a predefined, but disabled SMTP configuration called "SMTPConfig." You need to edit this configuration, specify your site-specific details, and then enable the configuration to get alert notification working properly. Note: You cannot delete "SMTPConfig" or add another SMTP configuration you can only edit the existing configuration. For more information on setting up user subscriptions to receive notifications for alert conditions, see "Editing the properties of a user account" on page In the top right, click. 2. Select SMTP Properties. 3. Review the configuration values and click to edit the properties as required: Property Name Name of the SMTP configuration. By default, the name of the existing SMTP configuration is "SMTPConfig," but you can change the name. SMTP configuration names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). Enabled Whether or not the SMTP configuration is enabled. By default, the initial default SMTP configuration is disabled because the SMTP configuration requires editing for your environment. When you enable the SMTP configuration, you must provide the required SMTP server and address properties. 28 Chapter 7: Managing the properties of a cluster

39 Viewing and editing the SMTP properties (Continued) Property SMTP Server (required) SMTP Server Port Display Name IP address or fully qualified host name of the SMTP server. For example: "smtp.mailserver.com" If specified as an IP address, use IPv4 notation, for example: The outgoing SMTP server port. The default port for a non-ssl configuration is 25. The default port for an SSL configuration is 465. The from display name associated with sent alert messages. For example: "admin." Display names can be from 1 to 255 characters and can contain any UTF-8 characters except ; (semicolon). Address (required) Security Enabled The from address associated with sent alert messages. For example: "WestCoastCluster@hds.com" Unicode characters are supported. Whether or not the client should use Secure Socket Layer (SSL) based security when sending . By default, Security Enabled is set to No. If you use SSL, you must provide a user name and password that will be passed to the service provider for authentication. User Name User name that is passed to the service provider for authentication. Required if Security Enabled is set to Yes. Password Password that is passed to the service provider for authentication. Required if Security Enabled is set to Yes. Passwords can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). of the SMTP configuration. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 7: Managing the properties of a cluster 29

40 Viewing and editing the SMTP properties Text input fields are provided for those properties that you can edit. 4. Click OK to save your changes, or Cancel to return to SMTP Properties without making changes. 30 Chapter 7: Managing the properties of a cluster

41 Managing IP addresses Managing IP addresses In addition to the virtual IP address that is associated with the cluster itself, external IP addresses from your public network are used by clients to access the data within the cluster. These IP addresses are specified as a range or pool when you initially configure your cluster network, and are then automatically and randomly assigned to each node in the cluster. IP addresses can move or float from the assigned node to a different node if the assigned node goes down or is taken off line for any reason, so there is no loss of service to the clients. Clients can access the cluster using any one of IP addresses associated with any one of the nodes. When directed to do so, IP addresses in the pool can also be assigned to VM instances. Important: You should maintain at least as many IP address as there are nodes in the cluster twice the number of IP addresses per node is recommended. Make sure you do not assign IP addresses to an HSP cluster that are already in use within your network. The IP Addresses page: Provides access to any alert conditions within the cluster Displays the list of IP addresses that have been added to the cluster Provides single-click access to all of the management operations you can perform on IP addresses: Click For help on this topic, see... "Adding an IP address" on page 34 "Editing the properties of an IP address" on page 37 "Deleting an IP address" on page 39 Select an IP address to view its properties. Chapter 7: Managing the properties of a cluster 31

42 Managing IP addresses Viewing the properties of an IP address You can edit some properties of an existing IP address resource. Important: You cannot edit the role of an IP address in the cluster. If you need to repurpose or change the role of an IP address, you will need to delete the IP address and then add it back specifying the new role you want the IP address to have. Specify the properties and corresponding values for any of the following properties that you want to edit: Property IP Address Type Select either: One IP Address IP Address Range Name Name of the IP address. IP address names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Enabled IP address's universally unique identifier (UUID). Whether or not the IP address is enabled. By default, the IP address is enabled. Specify either Yes or No. IP Address A valid external IP address that was allocated to the cluster when initially configured or that was added later to accommodate more storage/client connections. IP addresses are specified in IPv4 CIDR notation, for example: / Chapter 7: Managing the properties of a cluster

43 Managing IP addresses (Continued) Property Role Role assigned to the IP address in the configuration: cluster virtual IP address assigned to the cluster. node floating IP address assigned to nodes in the storage pool. vm IP address assigned to each of the virtual machine instances in the virtual machine pool. By default, role is set to node. of the IP address. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 7: Managing the properties of a cluster 33

44 Managing IP addresses Adding an IP address You can add IP addresses to the pool of floating IP addresses used for virtual machine instances and client connections to the cluster at any time. We recommend that you maintain at least as many IP address as there are nodes in the cluster for client connections twice the number of IP addresses as there are nodes is recommended. In you are going to run analytic applications in virtual machines on the cluster, you need to decide if you are going to allocate a pool of HSP managed IP addresses for the VM instances or you are going to use IP addresses managed from some external source (for example, DHCP). Even if you are going to manage IP addresses outside the cluster, you will need to create at least one virtual machine IP address so that you can run the a VM instance in sandbox mode and set up the address management. See "Working with a VM instance in sandbox mode" on page 74 If you do not specify a role, the IP address is assigned the role of node by default. 1. In the top right, click. 2. Select IP Addresses. 3. Click to add an IP address. 4. Enter the property values for the IP address or IP address range that you want to add: Property IP Address Type Select either: One IP Address IP Address Range 34 Chapter 7: Managing the properties of a cluster

45 Managing IP addresses (Continued) Property Name (required) Name of the IP address. Supply either a name for the IP address or enter "autogenerate." The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. IP address names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Enabled IP address's universally unique identifier (UUID). Whether or not the IP address is enabled. By default, the IP address is enabled. Specify either Yes or No. IP Address (required) A valid external IP address that was allocated to the cluster when initially configured or that was added later to accommodate more storage/client connections. IP addresses are specified in IPv4 CIDR notation, for example: /24. Role Role assigned to the IP address in the configuration: cluster virtual IP address assigned to the cluster. node floating IP address assigned to nodes in the storage pool. vm IP address assigned to each of the virtual machine instances in the virtual machine pool. By default, role is set to node. of the IP address. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 7: Managing the properties of a cluster 35

46 Managing IP addresses 5. Click OK to save your changes, or Cancel to return to the IP Addresses page without making changes. 36 Chapter 7: Managing the properties of a cluster

47 Managing IP addresses Editing the properties of an IP address You can edit some properties of an existing IP address resource. Important: You cannot edit the role of an IP address in the cluster. If you need to repurpose or change the role of an IP address, you will need to delete the IP address and then add it back specifying the new role you want the IP address to have. 1. In the top right, click. 2. Select IP Addresses. 3. Select an IP address that you want to edit. 4. Click to edit the IP address. 5. Enter the property values for the IP address that you want to change: Property Name Name of the IP address. IP address names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Enabled IP address's universally unique identifier (UUID). Whether or not the IP address is enabled. By default, the IP address is enabled. Specify either Yes or No. IP Address A valid external IP address that was allocated to the cluster when initially configured or that was added later to accommodate more storage/client connections. IP addresses are specified in IPv4 CIDR notation, for example: /24. Chapter 7: Managing the properties of a cluster 37

48 Managing IP addresses (Continued) Property Role Role assigned to the IP address in the configuration: cluster virtual IP address assigned to the cluster. node floating IP address assigned to nodes in the storage pool. vm IP address assigned to each of the virtual machine instances in the virtual machine pool. By default, role is set to node. of the IP address. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 6. Click OK to save your changes, or Cancel to return to the IP Addresses page without making changes. 38 Chapter 7: Managing the properties of a cluster

49 Managing IP addresses Deleting an IP address Deleting an IP address is removing the IP address from the HSP software s control and the IP address is no longer available for use inside the cluster. Important: You cannot delete the cluster's virtual IP address. 1. In the top right, click. 2. Select IP Addresses. 3. Select an IP address to delete. 4. Click to delete the IP address. 5. Click OK to confirm deletion. Chapter 7: Managing the properties of a cluster 39

50 Viewing and editing options Viewing and editing options 1. In the top right, click. 2. Select Options. 3. Review the configuration values and click to edit the options as required: Option Session Timeout Specify the session timeout in seconds. Maximum timeout value you can set is seconds ( days). GMT Offset Specify a negative number of the hours from Greenwich Mean Time. For example, -5 refers to a time zone that is five hours GMT. In Daylight Savings Time Whether or not the GMT offset specified is in a local that honors daylight savings time. In places not observing Daylight Saving Time the local UTC or GMT offset will remain the same year round. The Management Console will take of the offset adjustment if you set this to Yes. 4. Click OK to save your changes, or Cancel to return to Cluster Options without making changes. 40 Chapter 7: Managing the properties of a cluster

51 8 Managing virtual machine resources HSP provides the capability to host virtual machine images on the nodes of an HSP cluster to run applications on the same machines where the data for those applications resides. To ensure high availability, the HSP software monitors both node and VM health and attempts to failover and bring VM instances back online when a node goes down. You can use the Management Console to perform the administrative tasks needed to manage the virtual machine environment running on the cluster. The key to a successful virtual machine environment on an HSP cluster is not to over provision the VMs. The HSP software does provide some limits to help ensure VMs are not over provisioned in the cluster. These limits can deny that a VM be placed on a node that does not have enough memory. Here are important things to understand about VM provisioning: The memory limit for VMs is 2/3 of installed physical memory on a node. VM deployments are "all or nothing." If the HSP software cannot place all the requested VMs, the request fails with an error message. You can then choose between requesting a smaller number of VMs, requesting a smaller amount of memory for the VMs, or stopping other VMs to make room for the new ones. Placement for a VM failover is also subject to over provisioning limits. To successfully failover all of the VMs if a node goes down, you need to ensure you have enough unallocated memory on the rest of the nodes to handle the failover from the downed node. We recommend deploying larger VMs before smaller ones, so that multiple small VMs do not consume so many resources that there is not enough memory on any one node to place the larger VM, even though the cluster as a whole has enough memory. Chapter 8: Managing virtual machine resources 41

52 The Management Console lets you perform all the administrative tasks needed to manage the virtual machine resources in the cluster: Managing VM templates Managing VM instance groups Managing VM instances Managing VM instance snapshots Managing VM sizes Managing VM volumes 42 Chapter 8: Managing virtual machine resources

53 Managing VM templates Managing VM templates Virtual machine templates are used to install and manage the disk images used by the virtual machine software on HSP cluster nodes. Templates are either regular or golden. A golden template cannot be modified, which protects the contents of the image file. Golden templates let you distribute a base template that others can copy (clone) and modify to suit their needs. Say that you have a template that contains an image of the Linux operating system. This is considered a golden template because you want the applications deployed on your HSP cluster to run on the same version of Linux. Your applications team can clone this golden template and edit it to install Hadoop or other analytic application. Here is a summary of the steps required to get a virtual machine environment up and running on the Hyper Scale-Out Platform: 1. Copy the VM template to a shared out file system and add the VM template to HSP configuration so that it can be managed. 2. Clone the VM template if it is a protected golden template. (Cloning is not required for regular templates.) 3. Verify the properties associated with the template before instances and/or instance groups are deployed and make any necessary edits. If you have changed the base template considerably and you no longer want anyone to use an old template, you can delete a template. The VM Templates page: Provides access to any alert conditions within the cluster Displays the list of templates that have been added Provides single-click access to all of the management operations you can perform on VM templates: Chapter 8: Managing virtual machine resources 43

54 Managing VM templates Click For help on this topic, see... "Adding a VM template" on page 47 "Cloning a VM template" on page 50 "Editing the properties of a VM template" on page 49 "Deleting a VM template" on page 52 Select a VM template to view its properties. 44 Chapter 8: Managing virtual machine resources

55 Managing VM templates Viewing the properties of a VM template 1. On the VM Templates page, select the VM template for which you want to view properties. 2. To view the properties of a VM template: Click the name of the VM template if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name ID Is Golden Name of the VM template. VM template's universally unique identifier (UUID). Whether or not this template is a golden image. By default, Is Golden is set to to Yes. Image Path Path to the bootable VM disk image file. Image Format Format of the image file. The supported image formats are: auto-detect qcow2 raw By default, the software automatically detects whether the image format is raw or qcow2. Creation Time ModificationTime Time stamp of when the virtual machine template was created. Time stamp of when the virtual machine template was modified. The modification time stamp is different from creation time when the virtual machine template was modified in sandbox mode. Chapter 8: Managing virtual machine resources 45

56 Managing VM templates (Continued) Property Cluster Name Cluster ID Tenant Name Tenant ID Tags Name of the cluster to which the virtual machine template is associated. ID of the cluster to which the virtual machine template is associated. Name of the tenant who owns/manages the virtual machine template. ID of the tenant who owns/manages the virtual machine template. of the VM template. User-defined tags that can be used to organize and/or filter the response results. 46 Chapter 8: Managing virtual machine resources

57 Managing VM templates Adding a VM template Adding a VM template is the first HSP management API task you need to perform to get your virtual machine environment up and running. Prior to adding the VM template, you need to copy the bootable disk image file to a shared out HSP file system. A shared out file system is one that has both a share and access rule created and is available for client connections. 1. On the VM Templates page, click. 2. Enter the property values for the VM template that you want to add: Property Name (required) Name of the VM template. Supply either a name for the VM template or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. VM template names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Is Golden VM template's universally unique identifier (UUID). Whether or not this template is a golden image. By default, Is Golden is set to to Yes. Specify either Yes or No. Image Path (required) Path to the bootable VM disk image file. Specify the path in the following format: <share_name>:/full/path/to/image_file_name.img For example: vm_images:cirros_vm/cirros x86_64-disk.img Chapter 8: Managing virtual machine resources 47

58 Managing VM templates (Continued) Property Image Format Format of the image file. The supported image formats are: auto-detect qcow2 raw By default, if not specified when adding or editing a VM template, the software automatically detects whether the image format is raw or qcow2. of the VM template. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 3. Click OK to save your changes, or Cancel to return to the VM Templates page without making changes. 48 Chapter 8: Managing virtual machine resources

59 Managing VM templates Editing the properties of a VM template You can edit only a few properties associated with a template. 1. On the VM Templates page, click. 2. Select the VM template that you want to edit. 3. Enter the property values for the VM template that you want to change: Property Name Name of the VM template. VM template names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Image Format VM template's universally unique identifier (UUID). Format of the image file. The supported image formats are: auto-detect qcow2 raw By default, if not specified when adding or editing a VM template, the software automatically detects whether the image format is raw or qcow2. of the VM template. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Templates page without making changes. Chapter 8: Managing virtual machine resources 49

60 Managing VM templates Cloning a VM template The contents of a golden template are protected. However, you can modify or deploy a golden image by first copying or cloning that template. You can also clone a regular template if you want to make additional changes while preserving the original template. Cloning a template creates a new copy of the template. This includes creating a new directory, and creating a copy of the VM image file and the xml configuration file of the source template. This new template directory is created at the same path level as the source template. So if the golden template was located at <share>:/path/to/goldentemplate, the clone would be created as <share>:/path/to/uuid_of_new_template. The directory name of the cloned template is the UUID of the new template. 1. On the VM Templates page, select a VM template that you want to clone. 2. Click. 3. Enter the property values for the cloned VM template: Property Clone Name (required) Name of the cloned VM template. Supply either a name for the VM template or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. VM template clone names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). of the VM template clone. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 50 Chapter 8: Managing virtual machine resources

61 Managing VM templates 4. Click OK to save your changes, or Cancel to return to the VM Templates page without making changes. Chapter 8: Managing virtual machine resources 51

62 Managing VM templates Deleting a VM template Deleting a VM template will delete and remove the template from HSP administrative control. Use this command only if you no longer want add VM instance groups or instances using this VM template. 1. On the VM Templates page, select the VM template that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM template, by clicking OK. 52 Chapter 8: Managing virtual machine resources

63 Managing VM instance groups Managing VM instance groups VM instance groups provide a way for you to manage a group of instances that operate together. Instances that belong to an instance groups can have different size limit attributes and even use different VM templates. Complex analytic setups often require multiple products. A data analysis example, might require cooperating virtual machines for Hadoop, Cassandra, a relational DB, and Pentaho. Being able to group these virtual machines together in a virtual abstraction makes it easier to manage them as a unit. The VM Instance Groups page: Provides access to any alert conditions within the cluster Displays the list of VM instance groups that have been added to the cluster Provides single-click access to all of the management operations you can perform on VM instance groups: Click For help on this topic, see... "Adding a VM instance group" on page 55 "Editing the properties of a VM instance group" on page 62 "Viewing the list of VM instances associated with a VM instance group" on page 60 "Powering on the VM instances in a VM instance group" on page 63 "Shutting down the VM instances in a VM instance group" on page 64 "Deleting a VM instance group" on page 65 Select a VM instance group to view its properties. Chapter 8: Managing virtual machine resources 53

64 Managing VM instance groups Viewing the properties of a VM instance group 1. On the VM Instance Groups page, select the VM instance group for which you want to view properties. 2. To view the properties of a VM instance group: Click the name of the VM instance group if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name ID Run State Name of the VM instance group. VM instance group's universally unique identifier (UUID). Run state of the VM instance group: UP VM instance group is up. DOWN VM instance group is down. Num Instances Instance ID List Cluster Name Cluster ID Tenant Name Tenant ID Tags Number of instances in the VM instance group. Link to the VM instances in the VM instance group and their associated properties. Name of the cluster to which the VM instance group is associated. ID of the cluster to which the VM instance group is associated. Name of the tenant who owns/manages the VM instance group. ID of the tenant who owns/manages the VM instance group. of the VM instance group. User-defined tags that can be used to organize and/or filter the response results. 54 Chapter 8: Managing virtual machine resources

65 Managing VM instance groups Adding a VM instance group You can add a VM instance group to manage a group of virtual machine instances that operate together. HSP provides two approaches to setting up VM instance groups: You can add an "empty" instance group to which you can add multiple instances with the same sizing attributes (CPU and memory) or individual instances that have dissimilar sizing attributes. You can add multiple instances with the same sizing attributes (CPU and memory) at the same time that you create the instance group. The properties that you specify when adding a VM instance group depend on which approach you are taking. The property table below will guide you, so read the property descriptions carefully. No matter which approach you use, you can add instances to or delete instances from a VM instance group at any time. An IP address is required for each VM instance. If you do not have sufficient IP addresses, the deployment will fail. You need to decide if you are going to allocate these addresses from a pool of HSP-managed IP addresses or you are going to use IP addresses managed from some external source (for example, DHCP). See "Adding an IP address" on page 34 for information about adding HSP-managed IP addresses. If you are going to manage IP addresses outside the cluster, you may need to run the VM instance in sandbox mode and set up the address management. See "Working with a VM instance in sandbox mode" on page 74. If you are adding multiple instances at the same time, HSP names the instances for you. If you are adding a single instance, you have the option to name the instance. The default name format of a virtual machine instance depends on which kind of IP address that you use. If you are using the pool of HSP-managed IP addresses, the default instance name will include the IP address itself, something like vm If you use IP addresses from an external source, the default instance name will include the deploying template name, something like <template_name>-vm-1. In either case, the number at the end will increment for each instance deployed. Chapter 8: Managing virtual machine resources 55

66 Managing VM instance groups 1. On the VM Instance Groups page, click. 2. Enter the property values for the VM instance group that you want to add: Property Name (required) Name of the VM instance group. Supply either a name for the VM instance group or enter auto-generate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. Name is the only property you need to specify if you are adding an "empty" VM instance group and are not adding instances at the same time (Num Instances is not specified). VM instance group names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Run State VM instance group's universally unique identifier (UUID). Run state of the VM instance group: UP VM instance group is up. DOWN VM instance group is down. VM Template Name (required only if Num Instances is specified) VM Size (required only if Num Instances is specified) VM template to use for any VM instances added during VM instance group creation. VM size to use for any VM instances added during VM instance group creation. The current version of the software provides the following predefined sizes: jumbo If the vm-size is jumbo, then memory size is 16 GiB and the number of CPUs is 4. large If the vm-size is large, then memory size is 8 GiB and the number of CPUs is Chapter 8: Managing virtual machine resources

67 Managing VM instance groups (Continued) Property medium If the vm-size is medium, then memory size is 4 GiB and the number of CPUs is 2. small If the vm-size is small, then memory size is 2 GiB and the number of CPUs is 1. tiny If the vm-size is tiny, then memory size is 128 GiB and the number of CPUs is 1. You can also define a custom VM size to fit your needs. Supply either a name for the VM size or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. VM size names can be from 1 to 255 characters and can contain any UTF-8 characters. Num Instances Number of VM instances to add to the VM instance group during creation. If you want to add one or more instances at the same time you are adding the VM instance group, specify the number of instances that you want to add. If you add instances, you also need to specify the VM template and the VM size to use for the instances. Use Address Pool Whether or not IP assignments are to be made from the pool of IP addresses already designated for virtual machine instance use. By default, the HSP address pool designated for virtual machines is used to make IP address assignments. Specify either Yes or No. Tags of the VM instance group. User-defined tags that can be used to organize and/or filter the response results. 3. Click OK to save your changes, or Cancel to return to the VM Instance Groups page without making changes. Chapter 8: Managing virtual machine resources 57

68 Managing VM instance groups Adding VM instances to a VM instance group You can add one or more instances to an existing VM instance group. You can add instances with the same or different VM template and/or VM size attributes as other existing VM instances in the group. An IP address is required for each VM instance. If you do not have sufficient IP addresses, the deployment will fail. You need to decide if you are going to allocate these addresses from a pool of HSP-managed IP addresses or you are going to use IP addresses managed from some external source (for example, DHCP). See "Adding an IP address" on page 34 for information about adding HSP-managed IP addresses. If you are going to manage IP addresses outside the cluster, you may need to run the VM instance in sandbox mode and set up the address management. See "Working with a VM instance in sandbox mode" on page 74. If you are adding multiple instances at the same time, HSP names the instances for you. If you are adding a single instance, you have the option to name the instance. The default name format of a virtual machine instance depends on which kind of IP address that you use. If you are using the pool of HSP-managed IP addresses, the default instance name will include the IP address itself, something like vm If you use IP addresses from an external source, the default instance name will include the deploying template name, something like <template_name>-vm-1. In either case, the number at the end will increment for each instance deployed. 1. On the VM Instance Groups page, click. 2. On the VM Instances page, click. 3. Enter the property values for the VM instance or instances that you want to add: Property VM Template Name VM template to use for the VM instances being added. 58 Chapter 8: Managing virtual machine resources

69 Managing VM instance groups (Continued) Property Num Instances Number of VM instances to add to the VM instance group during creation. If you want to add one or more instances at the same time you are adding the VM instance group, specify the number of instances that you want to add. If you add instances, you also need to specify the VM template and the VM size to use for the instances. VM Size Name Use Address Pool Size of the VM instances being added. Whether or not IP assignments are to be made from the pool of IP addresses already designated for virtual machine instance use. By default, the HSP address pool for designated for virtual machines is used to make IP address assignments. Specify either Yes or No. Instance Name Name of the VM instance. of the VM instance. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Instance Groups page without making changes. Chapter 8: Managing virtual machine resources 59

70 Managing VM instance groups Viewing the list of VM instances associated with a VM instance group You can list the VM instances that are associated with a VM instance group. 1. On the VM Instance Groups page, select an VM instance group for which you want to view the list of VM instances. 2. Click. 60 Chapter 8: Managing virtual machine resources

71 Managing VM instance groups Deleting VM instances from a VM instance group Deleting a VM template will delete and remove the template from HSP administrative control. Use this command only if you no longer want add VM instance groups or instances using this VM template. 1. On the VM Instance Groups page, select the instance group from which you want to delete a VM instance. 2. Click. 3. On the VM Instances page, select the VM instance that you want to delete. 4. Click. 5. Confirm that you want to delete the selected VM instance, by clicking OK. Chapter 8: Managing virtual machine resources 61

72 Managing VM instance groups Editing the properties of a VM instance group You can edit only a few properties associated with a VM instance group. 1. On the VM Instance Groups page, click. 2. Select the VM instance group that you want to edit. 3. Enter the property values for the VM instance group that you want to change: Property Name Name of the VM instance group. VM instance group names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). of the VM instance group. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Instance Groups page without making changes. 62 Chapter 8: Managing virtual machine resources

73 Managing VM instance groups Powering on the VM instances in a VM instance group You can power on a group of VM instances that are associated with a VM instance group. 1. On the VM Instance Groups page, select the VM instance group for which you want to power on the instances. 2. Click. Chapter 8: Managing virtual machine resources 63

74 Managing VM instance groups Shutting down the VM instances in a VM instance group You can shut down a group of VM instances that are associated with a VM instance group. 1. On the VM Instance Groups page, select the VM instance group for which you want to shut down the instances. 2. Click. 64 Chapter 8: Managing virtual machine resources

75 Managing VM instance groups Deleting a VM instance group Deleting a VM instance group will remove the group and associated instances from HSP administrative control. Use this command only if you no longer want to manage instances using this VM instance group. Note: If the VM instance group you want to delete has running instances, you need to shut down those instances before you can delete the VM instance group. 1. On the VM Instance Groups page, select the VM instance group that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM instance group, by clicking OK. Chapter 8: Managing virtual machine resources 65

76 Managing VM instances Managing VM instances You can run VM instances on HSP nodes. The VM Instances page: Provides access to any alert conditions within the cluster Displays the list of VM instances that have been added to the cluster Provides single-click access to all of the management operations you can perform on VM instances: Click For help on this topic, see... "Adding VM instances" on page 70 "Editing the properties of a VM instance" on page 76 "Managing VM instance snapshots" on page 78 "Attaching a disk to a VM instance" on page 84 "Detaching a disk from a VM instance" on page 85 "Adding VM instances to a VM instance group" on page 58 "Deleting VM instances from a VM instance group" on page 61 "Migrating a VM instance " on page 77 "Powering on the VM instances in a VM instance group" on page 63 "Shutting down the VM instances in a VM instance group" on page 64 "Deleting a VM instance group" on the previous page Select a VM instance to view its properties. 66 Chapter 8: Managing virtual machine resources

77 Managing VM instances Viewing the properties of a VM instance 1. On the VM Instances page, select the VM instance for which you want to view properties. 2. To view the properties of a VM instance: Click the name of the VM instance if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name Name of the VM instance. The name you use also sets the hostname for the guest operating system. ID Run State VM instance's universally unique identifier (UUID). Run state of the VM instance: UP Instance is running. DOWN Instance or template deploying the instance has been shut down. IP address MAC address VNC address Memory Size Num CPUs Image Path IP address of the node on which the VM instance is running. MAC address of the node on which the VM instance is running. VNC address provides console or desktop access to the VM instance through a VNC viewer client. Amount of memory allowed for the VM instance resource. Number of CPUs allowed for the VM instance resource. Path to the bootable VM disk image file. Chapter 8: Managing virtual machine resources 67

78 Managing VM instances (Continued) Property Image Format Format of the VM image file. The supported image formats are: auto-detect qcow2 raw By default, the software automatically detects whether the image format is raw or qcow2. Is Sandbox A VM instance in sandbox mode is running a single instance of the actual template, not a copy of the template. You can edit the contents of a VM instance that is running in sandbox mode. By default, Is Sandbox is set to No. Disk VM Template Name VM Template ID VM Size Name VM Size ID VM Instance Group Name VM Instance Group ID Host Node Name Host Node ID Cluster Name Cluster ID Tenant Name Tenant ID The list of local disks attached to the VM instance. Name of the VM template to which the VM instance is associated. ID of the VM template to which the VM instance is associated. Name of the VM template to which the VM instance is associated. ID of the VM template to which the VM instance is associated. Name of the VM instance group to which the VM instance is associated. ID of the VM instance group to which the VM instance is associated. Host name of the node on which the VM instance is running. Host ID of the node on which the VM instance is running. Name of the cluster to which the VM instance is associated. ID of the cluster to which the VM instance is associated. Name of the tenant who owns/manages the VM instance. ID of the tenant who owns/manages the VM instance. 68 Chapter 8: Managing virtual machine resources

79 Managing VM instances (Continued) Property Tags of the VM instance. User-defined tags that can be used to organize and/or filter the response results. Chapter 8: Managing virtual machine resources 69

80 Managing VM instances Adding VM instances You can add a VM instance to an existing VM instance group or you can add a standalone instance. Adding an VM instance starts the guest operating system for the instance. An IP address is required for each VM instance. If you do not have sufficient IP addresses, the deployment will fail. You need to decide if you are going to allocate these addresses from a pool of HSP-managed IP addresses or you are going to use IP addresses managed from some external source (for example, DHCP). See "Adding an IP address" on page 34 for information about adding HSP-managed IP addresses. If you are going to manage IP addresses outside the cluster, you may need to run the VM instance in sandbox mode and set up the address management. See "Working with a VM instance in sandbox mode" on page 74. If you are adding multiple instances at the same time, HSP names the instances for you. If you are adding a single instance, you have the option to name the instance. The default name format of a virtual machine instance depends on which kind of IP address that you use. If you are using the pool of HSP-managed IP addresses, the default instance name will include the IP address itself, something like vm If you use IP addresses from an external source, the default instance name will include the deploying template name, something like <template_name>-vm-1. In either case, the number at the end will increment for each instance deployed. 1. On the VM Instances page, click. 70 Chapter 8: Managing virtual machine resources

81 Managing VM instances 2. Enter the property values for the VM instance or instances that you want to add: Property Name (required) Name of the VM instance. The name you use also sets the hostname for the guest operating system. Follow the RFC 1123 host name guidelines. VM instance names can be from 1 to 255 characters and can only contain: letters A-Z and a-z digits 0-9 hyphens (-) VM Template Name VM template to use for the VM instances being added. (required) VM Size Name Size of the VM instances being added. (required) Use Address Pool Whether or not IP assignments are to be made from the pool of IP addresses already designated for virtual machine instance use. By default, the HSP address pool for designated for virtual machines is used to make IP address assignments. Specify either Yes or No. Is Sandbox A VM instance in sandbox mode is running a single instance of the VM template image. You can edit the contents of a VM instance that is running in sandbox mode. By default, Is Sandbox is set to No. Specify either Yes or No. Tenant Name Disk Tenant who owns/manages the VM instance. The list of local disks to which the VM instance has been attached. Chapter 8: Managing virtual machine resources 71

82 Managing VM instances (Continued) Property of the VM instance. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 3. Click OK to save your changes, or Cancel to return to the VM Instances page without making changes. 72 Chapter 8: Managing virtual machine resources

83 Managing VM instances Powering on a VM instance You can power on a particular VM instance. 1. On the VM Instances page, select the VM instance that you want to power on. 2. Click. Chapter 8: Managing virtual machine resources 73

84 Managing VM instances Working with a VM instance in sandbox mode You can edit the contents of a copy of VM template image using a VM instance running in sandbox mode. You need to create an instance, specifying the sandbox property. Once you complete your edits to the image, you can delete the instance. Note: You cannot edit a golden template you can only edit a cloned copy of a golden template. You also cannot edit the VM image if that template is being used for other running VM instances. 1. On the VM Instances page, click. 2. Enter the property values for the VM instance or instances that you want to add, ensuring that you set the Is Sandbox property to Yes: Property Name (required) Name of the VM instance. The name you use also sets the hostname for the guest operating system. Follow the RFC 1123 host name guidelines. VM instance names can be from 1 to 255 characters and can only contain: letters A-Z and a-z digits 0-9 hyphens (-) VM Template Name VM template to use for the VM instances being added. (required) VM Size Name Size of the VM instances being added. (required) 74 Chapter 8: Managing virtual machine resources

85 Managing VM instances (Continued) Property Use Address Pool Whether or not IP assignments are to be made from the pool of IP addresses already designated for virtual machine instance use. By default, the HSP address pool for designated for virtual machines is used to make IP address assignments. Specify either Yes or No. Is Sandbox A VM instance in sandbox mode is running a single instance of the VM template image. You can edit the contents of a VM instance that is running in sandbox mode. By default, Is Sandbox is set to No. Specify either Yes or No. Tenant Name Disk Tenant who owns/manages the VM instance. The list of local disks to which the VM instance has been attached. of the VM instance. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 8: Managing virtual machine resources 75

86 Managing VM instances Editing the properties of a VM instance You can edit only a few properties associated with a virtual machine instance. 1. On the VM Instances page, click. 2. Select the VM instance that you want to edit. 3. Enter the property values for the VM instance that you want to change: Property Name Name of the VM instance. The name you use also sets the hostname for the guest operating system. Follow the RFC 1123 host name guidelines. VM instance names can be from 1 to 255 characters and can only contain: letters A-Z and a-z digits 0-9 hyphens (-) of the VM instance. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Instances page without making changes. 76 Chapter 8: Managing virtual machine resources

87 Managing VM instances Migrating a VM instance If you need to shut down or retire a node, migration lets you to shutdown the running VM instance on one node and restart it on another node you specify. You can also use migration to manually load balance virtual machines when you are not satisfied with the automatic round robin placement scheme made the HSP software. Important: You cannot migrate VMs that have local virtual machine disks attached. Note: VM instance migration is not "live migration." The running VM instance is shut down and restarted as if the power button was cycled. We therefore recommend that you gracefully shut down any applications running on the virtual machine before migration. 1. On the VM Instances page, select the VM instance that you want to migrate. 2. Click. 3. Enter the property values for the migration: Node (required) Property Force Shutdown Specify the node to which you want to migrate the VM instance. Whether or not to force the shutdown of running instances. Specify the force option if you have not or were not able to shut down the instances gracefully. 4. Click OK to save your changes, or Cancel to return to the VM Instances page without making changes. Chapter 8: Managing virtual machine resources 77

88 Managing VM instance snapshots Managing VM instance snapshots A virtual machine snapshot is a point-in-time picture of the state of a running VM instance when the snapshot is taken. The snapshot preserves both the disk state and memory state of the VM. Consider taking a VM snapshot periodically so that you have a stable state to revert to if something goes wrong. A snapshot would let you return the deployed and running VM instance back to a known good point. You can only snapshot instances deployed from a template with a qcow2 image format, raw image formats are not supported. Snapshots are kept in the same qcow2 file from which the instance is running. You can take as many snapshots as you have space for in your instance. So if you created a virtual machine with 50 GB of disk space, you can take snapshots until you run you run out of space. Snapshots are only retained for as long as the VM instance is running. Once a VM instance has been shutdown, all snapshots are deleted. The VM Snapshots page: Provides access to any alert conditions within the cluster Displays the list of VM snapshots that have been taken of a VM instance Provides single-click access to all of the management operations you can perform on VM snapshots: Click For help on this topic, see... "Adding a VM snapshot" on page 80 "Editing a VM snapshot" on page 81 "Reverting a VM snapshot" on page 82 "Deleting a VM snapshot" on page 83 Select a VM snapshot to view its properties. 78 Chapter 8: Managing virtual machine resources

89 Managing VM instance snapshots Listing the VM snapshots for a VM instance 1. In VM Instances page, select a VM instance for which you want to view snapshots. 2. Click. A list of snapshots is returned. Chapter 8: Managing virtual machine resources 79

90 Managing VM instance snapshots Adding a VM snapshot Adds a snapshot of the specified VM instance. Note: You can only snapshot instances deployed from a template with a qcow2 image format, raw image formats are not supported. 1. On the VM Instances page, select the VM instance that you want to snapshot. 2. Click. 3. On the VM Snapshots page, click. 4. Enter the property values for the VM snapshots that you want to add: Property Type Name string Name of the VM snapshot. Names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). string of the VM snapshot. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags string User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 5. Click OK to save your changes, or Cancel to return to the VM Snapshots page without making changes. 80 Chapter 8: Managing virtual machine resources

91 Managing VM instance snapshots Editing a VM snapshot You can edit a snapshot of the specified VM instance. 1. On the VM Snapshots page, click. 2. Select the VM snapshot that you want to edit. 3. Enter the property values for the VM snapshot that you want to change: Property Type Name string Name of the VM snapshot. Names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). string of the VM snapshot. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags string User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Snapshots page without making changes. Chapter 8: Managing virtual machine resources 81

92 Managing VM instance snapshots Reverting a VM snapshot Reverts a snapshot of the specified VM instance back to the specified snapshot. 1. On the VM Instances page, select the VM instance that you want to revert snapshot. 2. Click. 3. On the VM Snapshots page, click. 4. Provide the name of the snapshot to which you want to revert. Property Type Name string Name of the VM snapshot. Names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). 82 Chapter 8: Managing virtual machine resources

93 Managing VM instance snapshots Deleting a VM snapshot You can delete a snapshot of the specified VM instance. 1. On the VM Snapshots page, select the VM snapshot that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM snapshot, by clicking OK. Chapter 8: Managing virtual machine resources 83

94 Managing VM instance snapshots Attaching a disk to a VM instance You can associate one or more existing VM volumes with a local HSP disk and attach the disk to a VM instance. You need to edit the role of the disk from cluster to virtual machine prior to adding volumes and associating them with a disk. Note: The VM instance may need to be rebooted before the VM disk is seen by the guest operating system. 1. On the VM Instances page, select the VM instance for which you want to attach a disk. 2. Click. 3. Enter the property values for the VM volumes associated with the disk you want to attach: VM Disk Property A comma-separated list of VM volumes to be associated with the disk and attached to the VM instance. In this version of the software, only local disks are supported. Specify the disk as one or more VM volumes: vm-volume-id=<id>,vm-volume-id=<id> 84 Chapter 8: Managing virtual machine resources

95 Managing VM instance snapshots Detaching a disk from a VM instance You can detach one or more existing VM volumes from a local HSP disk that is attached to VM instance. 1. On the VM Instances page, select the VM instance for which you want to detach a disk. 2. Click. Chapter 8: Managing virtual machine resources 85

96 Managing VM instance snapshots Rebooting a VM instance You can reboot a VM instance. 1. On the VM Instances page, select the VM instance that you want to reboot. 2. Click. 3. Confirm that you want to reboot the selected VM instance, by clicking OK. 86 Chapter 8: Managing virtual machine resources

97 Managing VM instance snapshots Shutting down a VM instance You can shut down a particular VM instance. 1. On the VM Instances page, select the VM instance that you want to shut down. 2. Click. 3. Confirm that you want to shut down the selected node, by clicking OK. Chapter 8: Managing virtual machine resources 87

98 Managing VM instance snapshots Deleting a VM instance Deleting a VM instance will shutdown the guest operating system and remove the instance from HSP administrative control. Use this command only if you no longer want to run the VM instance in an HSP cluster. 1. On the VM Instances page, select the VM instance that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM instance, by clicking OK. 88 Chapter 8: Managing virtual machine resources

99 Managing VM sizes Managing VM sizes The VM size resource defines the maximum CPU and memory resources available to a VM instance. For best VM performance and cluster performance, we recommend taking care not to over or under specify these resources for your application. You can specify one of the HSP pre-defined VM sizes or you can define a custom VM size. The VM Sizes page: Provides access to any alert conditions within the cluster Displays the list of VM sizes that have been added Provides single-click access to all of the management operations you can perform on VM sizes: Click For help on this topic, see... "Adding a VM size" on page 92 "Editing the properties of a VM size" on page 93 "Deleting a VM size" on page 94 Select a VM size to view its properties. Chapter 8: Managing virtual machine resources 89

100 Managing VM sizes Viewing the properties of a VM size 1. On the VM Sizes page, select the VM size for which you want to view properties. 2. To view the properties of a VM size: Click the name of the VM size if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name (required) Name of the VM size. The current version of the software provides the following pre-defined sizes: jumbo If the vm-size is jumbo, then memory size is 16 GiB and the number of CPUs is 4. large If the vm-size is large, then memory size is 8 GiB and the number of CPUs is 2. medium If the vm-size is medium, then memory size is 4 GiB and the number of CPUs is 2. small If the vm-size is small, then memory size is 2 GiB and the number of CPUs is 1. tiny If the vm-size is tiny, then memory size is 128 GiB and the number of CPUs is 1. You can also define a custom VM size to fit your needs. Supply either a name for the VM size or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. VM size names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID VM size's universally unique identifier (UUID). 90 Chapter 8: Managing virtual machine resources

101 Managing VM sizes (Continued) Property Memory Size (required) Num CPUs (required) Cluster Name Cluster ID Tenant Name Tenant ID Amount of memory allowed for a VM instance resource. You can specify a memory size from 512 MiB to 128 GiB. Number of CPUs allowed for a VM instance resource. You can specify from 1 to 8 CPUs. Name of the cluster to which the VM size is associated. ID of the cluster to which the VM size is associated. Name of the tenant associated with the VM size. ID of the tenant associated with the VM size. of the VM size. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). Chapter 8: Managing virtual machine resources 91

102 Managing VM sizes Adding a VM size If the predefined HSP VM sizes do not suit your needs, you can add a VM size that can customize the CPU and memory resources you require. 1. On the VM Sizes page, click. 2. Enter the property values for the VM size that you want to add: Property Name (required) Name of the VM size. Supply either a name for the VM size or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. VM size names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Memory Size (required) Num CPUs (required) VM size's universally unique identifier (UUID). Amount of memory allowed for a VM instance resource. You can specify a memory size from 512 MiB to 128 GiB. Number of CPUs allowed for a VM instance resource. You can specify from 1 to 8 CPUs. of the VM size. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 3. Click OK to save your changes, or Cancel to return to the VM Sizes page without making changes. 92 Chapter 8: Managing virtual machine resources

103 Managing VM sizes Editing the properties of a VM size You can edit the properties any of the VM sizes, including the predefined HSP VM sizes. 1. On the VM Sizes page, click. 2. Select the VM size that you want to edit. 3. Enter the property values for the VM size that you want to change: Property Name Name of the VM size. VM size names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Memory Size VM size's universally unique identifier (UUID). Amount of memory allowed for a VM instance resource. You can specify a memory size from 512 MiB to 128 GiB. Num CPUs Number of CPUs allowed for a VM instance resource. You can specify from 1 to 8 CPUs. of the VM size. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Sizes page without making changes. Chapter 8: Managing virtual machine resources 93

104 Managing VM sizes Deleting a VM size You can delete any of the VM sizes, including the predefined HSP VM sizes. 1. On the VM Sizes page, select the VM size that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM size, by clicking OK. 94 Chapter 8: Managing virtual machine resources

105 Managing VM volumes Managing VM volumes Some high availability applications, like Hadoop and NoSQL, need to manage their own storage and do not require the built-in high availability features of HSP. VM volumes let you manage the local volume space for virtual machines running in an HSP cluster. Having this dedicated storage improves the performance of applications that do a lot of small writes. VM volumes are created on an HSP disk and then attached to VM instances. Note: Only a cluster administrator can add, edit, or delete VM volumes. The VM Volumes page: Provides access to any alert conditions within the cluster Displays the list of VM volumes that have been added Provides single-click access to all of the management operations you can perform on VM volumes: Click For help on this topic, see... "Adding VM volumes" on page 98 "Editing the properties of a VM volume" on page 99 "Editing the properties of a VM volume" on page 99 Select a VM volume to view its properties. Chapter 8: Managing virtual machine resources 95

106 Managing VM volumes Viewing the properties of a VM volume 1. On the VM Volumes page, select the VM volume for which you want to view properties. 2. To view the properties of a VM volume: Click the name of the VM volume if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name ID Run State Name of the VM volume. VM volume's universally unique identifier (UUID). Run state of the VM volume: UP VM Volume has been added and available for use. ERROR The underlying disk to which the VM volume is associated has failed. Capacity VM InstanceName VM Instance ID Disk Name Disk ID Source Devname Target Devname Cluster Name Cluster ID Tenant Name Tenant Name Tenant ID Capacity of the VM Volume. Name of the VM instance to which the VM volume is associated. ID of the VM instance to which the VM volume is associated. The name of the disk to which the VM Volume is associated. The ID of the disk to which the VM Volume is associated. The LVM2 volume device name on the host HSP node. Block device name inside the guest VM. Name of the cluster to which the VM volume is associated. ID of the cluster to which the VM volume is associated. The tenant who owns/manages the VM volume. Name of the tenant who owns/manages the VM volume. ID of the tenant who owns/manages the VM volume. 96 Chapter 8: Managing virtual machine resources

107 Managing VM volumes (Continued) Property Tags of the VM volume. User-defined tags that can be used to organize and/or filter the response results. Chapter 8: Managing virtual machine resources 97

108 Managing VM volumes Adding VM volumes You can add a VM volume that will let you dedicate space for a virtual machine instance. After you have added one or more VM volumes you need to specify the disk used for the volumes and attach the disk to a VM instance. See "Attaching a disk to a VM instance" on page 84 Important: You need to edit the role of the disk from cluster to virtual machine prior to adding volumes and associating them with a disk. 1. On the VM Volumes page, click. 2. Enter the property values for the VM volume that you want to add: Property Name ID Disk Name Disk ID Source Devname Target Devname Cluster Name Cluster ID Tenant Name Tenant Name Tags Name of the VM volume. VM volume's universally unique identifier (UUID). The name of the disk to which the VM Volume is associated. The ID of the disk to which the VM Volume is associated. The LVM2 volume device name on the host HSP node. Block device name inside the guest VM. Name of the cluster to which the VM volume is associated. ID of the cluster to which the VM volume is associated. The tenant who owns/manages the VM volume. Name of the tenant who owns/manages the VM volume. of the VM volume. User-defined tags that can be used to organize and/or filter the response results. 3. Click OK to save your changes, or Cancel to return to the VM Volumes page without making changes. 98 Chapter 8: Managing virtual machine resources

109 Managing VM volumes Editing the properties of a VM volume You can edit only a few properties associated with a VM volume. 1. On the VM Volumes page, click. 2. Select the VM volume that you want to edit. 3. Enter the property values for the VM volume that you want to change: Property Name ID Disk Name Disk ID Source Devname Target Devname Cluster Name Cluster ID Tenant Name Tenant Name Name of the VM volume. VM volume's universally unique identifier (UUID). The name of the disk to which the VM Volume is associated. The ID of the disk to which the VM Volume is associated. The LVM2 volume device name on the host HSP node. Block device name inside the guest VM. Name of the cluster to which the VM volume is associated. ID of the cluster to which the VM volume is associated. The tenant who owns/manages the VM volume. Name of the tenant who owns/manages the VM volume. of the VM volume. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the VM Volumes page without making changes. Chapter 8: Managing virtual machine resources 99

110 Managing VM volumes Deleting a VM volume Deleting a VM volume will remove the volume from HSP administrative control. Use this command only if you no longer want to manage the defined volume space in an HSP cluster. 1. On the VM Volumes page, select the VM volume that you want to delete. 2. Click. 3. Confirm that you want to delete the selected VM volume, by clicking OK. 100 Chapter 8: Managing virtual machine resources

111 9 Managing cluster resources The Management Console lets you perform all the administrative tasks needed to manage the file storage, hardware, and user account resources in the cluster: Managing the properties of a cluster Managing IP addresses Managing tenants Managing file systems Managing shares Managing racks Managing nodes Managing disks Managing user accounts Chapter 9: Managing cluster resources 101

112 Managing tenants Managing tenants Tenants own and manage file systems and file system-related resources, as well as virtual machine resources. While multiple tenants share the HSP hardware, storage, and software capabilities within a cluster, they do not share or see each other's data or virtual machine applications. Defining a tenant resource is not required if you simply want to manage HSP cluster resources as a single entity with shared data. Multitenancyrelated fields will be present in certain --json outputs (for example tenantname and tenant-id), but these fields will be empty. The first step in implementing multitenancy in an HSP cluster, is to create one or more tenants. A tenant typically corresponds to an organization such as a company or a division or department within a company. A tenant can also correspond to an individual person. A cluster administrator creates tenants and the HSP software automatically creates an administrative user account "tenant-name*admin" for each tenant when that tenant is created. Tenant administrators have management access to the following tenant resources it manages: alert condition event logs file systems shares roles users VM templates VM instance groups VM instances The cluster administrator is a privileged user and also has access to all tenant resources and can perform all management operations. See "Managing user accounts" on page 172 for more information specific to creating tenant users and the roles and privileges that can be assigned them. The Tenants page: Provides access to any alert conditions within the cluster 102 Chapter 9: Managing cluster resources

113 Managing tenants Displays the list of tenants that have been added to the cluster Provides single-click access to all of the management operations you can perform on tenants: Click For help on this topic, see... "Adding a tenant" on page 106 "Editing the properties of a tenant" on page 107 "Deleting a tenant" on page 108 Select a tenant to view its properties. Chapter 9: Managing cluster resources 103

114 Managing tenants Viewing the properties of a tenant 1. On the Tenants page, select the tenant for which you want to view properties. 2. To view the properties of a tenant: Click the name of the tenant if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name ID Enabled Name of the tenant. Tenant's universally unique identifier (UUID). Whether or not the tenant is enabled. By default, the tenant is enabled. File System Space HWM High water mark percentage set by default for a new file system added to a cluster. You can override this percentage at the file system level. When the remaining allocated file system space reaches the high water mark, an alert is raised until the file system space consumption falls below the low water mark percentage. By default, file system space high water mark is set to 90. File System Space LWM Low water mark percentage set by default for a new file system added to a cluster. Existing file system space alerts are automatically cleared once the low water mark setting is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. By default, the file system space low water mark is set to 85. Cluster Name Cluster ID Name of the cluster to which the tenant is associated. ID of the cluster to which the tenant is associated. 104 Chapter 9: Managing cluster resources

115 Managing tenants (Continued) Property Tags of the tenant. User-defined tags that can be used to organize and/or filter the response results. Chapter 9: Managing cluster resources 105

116 Managing tenants Adding a tenant HSP cluster administrators can create tenants. When you add a new tenant and the HSP software automatically creates the administrative user account "tenant-name*admin" to manage that tenant. 1. On the Tenants page, click. 2. Enter the property values for the tenant that you want to add: Property Name (required) Name of the tenant. Supply either a name for the tenant or enter "autogenerate." The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. Tenant names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk) and ; (semicolon). ID Enabled Tenant's universally unique identifier (UUID). Whether or not the tenant is enabled. By default, the tenant is enabled. Specify either Yes or No. of the tenant. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 3. Click OK to save your changes, or Cancel to return to the Tenants page without making changes. 106 Chapter 9: Managing cluster resources

117 Managing tenants Editing the properties of a tenant You can edit the description and tags associated with the tenant, as well as enabling or disabling the tenant. Important: If you need to change the name of a tenant, you will need to add a new tenant and then delete the old tenant. 1. On the Tenants page, click. 2. Select the tenant that you want to edit. 3. Enter the property values for the tenant that you want to change: Property ID Enabled Tenant's universally unique identifier (UUID). Whether or not the tenant is enabled. By default, the tenant is enabled. Specify either Yes or No. of the tenant. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the Tenants page without making changes. Chapter 9: Managing cluster resources 107

118 Managing tenants Deleting a tenant Deleting a tenant also deletes the associated tenant administrator. 1. On the Tenants page, select the tenant that you want to delete. 2. Click. 3. Confirm that you want to delete the selected tenant, by clicking OK. 108 Chapter 9: Managing cluster resources

119 Managing file systems Managing file systems A file system is a namespace in which files are created and accessed. The HSP concept of a file system differs from the traditional concept of a file system in the following ways: Each file exists in only one file system, but instead of being stored on individual storage devices, the files are stored on multiple disks and spread across multiple nodes. You manage file system space usage by specifying a quota (size limit) for each file system. HSP monitors and enforces these quotas and issues alerts when you are approaching quota limits. You can easily change the quota for a file system without any interruption in service. Space is only allocated as files are written, so you can create file system quotas that are much larger than the space that is currently available on physical disks. The quota for a file system is also not a guaranty file systems share and can only grow to the actual storage space available within the cluster. Keep in mind that in a distributed file system, reaching the maximum limit of a quota cannot happen with the same preciseness possible in a nondistributed file system. By the time space tracking information has been extracted and synchronized across the nodes, some amount of data may be written that exceeds the quota. By default, all file system data writes include a companion checksum write to verify data integrity. When data is read, a checksum is calculated from the on-disk data and compared to the stored checksum. If the computed checksum does not match the stored checksum, HSP takes the appropriate measures to protect against data corruption, based on the data protection policy you specified for the file system. Important: File system space tracking and high and low water mark calculations for alerts are based on disk space statistics reported by the HSP storage layer. For performance reasons, these statistics are reported at set intervals, rather than dynamically. You may, therefore, see a slight lag between what is displayed for size and use statistics in HSP and what a more dynamic command like df shows. Chapter 9: Managing cluster resources 109

120 Managing file systems The File Systems page: Provides access to any alert conditions within the cluster Displays the list of file systems that have been added to the cluster Provides single-click access to all of the management operations you can perform on file systems: Click For help on this topic, see... "Adding a file system" on page 116 "Editing the properties of a file system" on page 122 "Enabling or disabling a file system" on page 121 "Checking and repairing a file system" on page 126 "Managing shares" on page 128 "Deleting a file system" on page 127 Select a file system to view its properties. 110 Chapter 9: Managing cluster resources

121 Managing file systems Viewing the properties of a file system You can edit the properties of an existing file system, including the quota limit. Quota changes can be made at any time without any client interruption in service. 1. On the File Systems page, select the file system for which you want to view properties. 2. To view the properties of a file system: Click the name of the file system if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name ID Run State Name of the file system. File system's universally unique identifier (UUID). Run state of the file system: UP File system is fully functional, and data is available. DOWN File system is offline and the data is not available. An administrator might have disabled the file system. IMPAIRED File system has reported errors. We recommend running fsck to check and repair the file system. A warning severity alert is raised. ERROR File system is down due to errors that prevent it from staying online. We recommend running fsck to check and repair the file system. Customer support intervention may be required to troubleshoot the problem if fsck is not successful in repairing the file system. An error severity alert is raised. See "Checking and repairing a file system" on page 1 for more information about running the fsck utility on file systems that have a run state of IMPAIRED or ERROR. Chapter 9: Managing cluster resources 111

122 Managing file systems (Continued) Property Enabled Whether or not the file system is enabled. By default, the file system is enabled. Shared Out Whether or not the file system is shared out and available for client connections. A file system must have both a share and an access rule associated with it before clients can connect to it and therefore be considered shared out. 112 Chapter 9: Managing cluster resources

123 Managing file systems (Continued) Property Data Protection Data protection policy set for a file system. By default, the data protection policy is set to raid1_ 3instance_large_io. The current version of the software supports the following values: raid1_3instance_large_io Mirrored with 3 copies or instances of the data, tuned for supporting large streaming writes. This is the ideal for most Big Data Analytics style workloads or general purpose Data Lake. raid1_3instance_small_io: Mirrored with 3 copies or instances of the data, tuned for supporting small random writes. This is ideal for hosting virtual machine images and the best choice for relational databases. raid0_1instance_small_io: A single instance of the data is kept, data is not highly available. It is tuned for supporting small random writes. This is ideal for scratch space where low storage consumption is desired and workloads may be unpredictable, but losing data due to hardware failures would not be catastrophic. HSP automatically ensures that multiple instances of any given file are stored on multiple nodes based on the data protection policy that you specify for a file system. If a node or disk that contains a particular file fails, HSP automatically finds another instance and forwards the client request to another node for one of the file instances. Keep in mind, however, the more copies of each file you retain, the greater the storage consumption, so you need to determine the data protection/space consumption balance that makes sense for your file system. Creation Time Time stamp of when the file system was created. Chapter 9: Managing cluster resources 113

124 Managing file systems (Continued) Property Space HWM Space LWM encryption-algorithm File system space usage high water mark percentage. A high water mark percentage set at the file system level overrides the cluster-level property File System Space HWM that is set by default for new file systems added to the cluster. File system space usage low water mark percentage. A low water mark percentage set at the file system level overrides the cluster-level property File System Space LWM that is set by default for new file systems added to the cluster. AES128 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 128-bit keys. AES256 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 256-bit keys. This encryption algorithm offers stronger cryptographic protection than AES128. none - no encryption. Quota An administrator-defined maximum file system size limit. HSP monitors and enforces file system quotas and issues alerts when you are approaching quota limits. Specify a value and unit from 128 GiB to 6 PiB. Used Capacity Free Capacity used-inodesused Inodes Free Inodes Record Access Time fsck Action Running fsck Job ID fsck Job Details Used capacity in the file system, specified in bytes. Free capacity in file system, specified in bytes Number of used inodes (files+directories+symbolic links) on the file system. Number of free inodes (files+directories+symbolic links) on the file system. Whether or not the file system should track the last time a file was accessed. Whether or not an active fsck is running on the file system. ID of the job associated with the file system fsck. Job details of an active fsck. 114 Chapter 9: Managing cluster resources

125 Managing file systems (Continued) Property fsck Percent Complete Cluster Name HSP Internal Progress of an active fsck, expressed as a percentage. Name of the cluster to which the file system is associated. Whether or not the file system is an HSP internal file system. Internal file systems are created and maintained by the HSP software. These file systems cannot be deleted or renamed. Cluster ID Tenant Name Tenant ID Tags ID of the cluster to which the file system is associated. Name of the tenant who owns/manages the file system. ID of the tenant who owns/manages the file system. of the file system. User-defined tags that can be used to organize and/or filter the response results. Chapter 9: Managing cluster resources 115

126 Managing file systems Adding a file system Create files systems in an organizational manner that makes sense for your company and the files you are storing, for example, by business organization/unit, by file type, by project, and so on. You can add up to 1024 virtual file systems in each cluster. Because the size of a file system is not limited by the underlying storage, file system size limits or quotas are purely space management limits that you want to set for organizational categories you choose. Physical storage is allocated only as files are written to a file system, so you do not need to allocate a physical storage pool of equal size. Another thing to keep in mind, is that data protection policies (how many instances of each file in that file system you want to store) are set at the file system level. Data protection policies then, may play also role in your file system organization creation scheme. The root permissions for a new file system are set to 777 by default. Client access permissions are then granted to the file system by adding an NFS share and by any specific access control rules you create for that share. See "Adding a share" on page 1 for more information. There is a cluster-level property named File System Auto Access that controls whether or not you will need to manually create the share and access rule required to share out the newly added file system or the whether the HSP software will automatically create these for you. By default, File System Auto Access is set to True and the file system is shared out. When you allow the software to automatically share out the file system, HSP adds a share with the same name as the file system and adds an access rule called "DefaultAccessRule" that provides read/write access to the share for all client hosts. You can add more restrictive access rules to the share, as well as edit or delete this automatically created access rule. If at the cluster level, however, File System Auto Access, you can mark Auto Access at file system creation time to override the cluster setting and automatically create the share and default access rule. Otherwise, you will need to manually create the required share and access rule to share out the file system. 1. On the File Systems page, click. 116 Chapter 9: Managing cluster resources

127 Managing file systems 2. Enter the property values for the file system that you want to add: Property Name (required) Name of the file system. Supply either a name for the file system or enter autogenerate. The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. The name of the file system you provide is also the mount point for the file system, so follow standard UNIX file system naming conventions and rules. File system names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk), / (slash), and ; (semicolon). ID Enabled File system's universally unique identifier (UUID). Whether or not the file system is enabled. By default, the file system is enabled. Specify either Yes or No. Chapter 9: Managing cluster resources 117

128 Managing file systems (Continued) Property Data Protection Data protection policy set for a file system. By default, the data protection policy is set to raid1_ 3instance_large_io. The current version of the software supports the following values: raid1_3instance_large_io Mirrored with 3 copies or instances of the data, tuned for supporting large streaming writes. This is the ideal for most Big Data Analytics style workloads or general purpose Data Lake. raid1_3instance_small_io: Mirrored with 3 copies or instances of the data, tuned for supporting small random writes. This is ideal for hosting virtual machine images and the best choice for relational databases. raid0_1instance_small_io: A single instance of the data is kept, data is not highly available. It is tuned for supporting small random writes. This is ideal for scratch space where low storage consumption is desired and workloads may be unpredictable, but losing data due to hardware failures would not be catastrophic. HSP automatically ensures that multiple instances of any given file are stored on multiple nodes based on the data protection policy that you specify for a file system. If a node or disk that contains a particular file fails, HSP automatically finds another instance and forwards the client request to another node for one of the file instances. Keep in mind, however, the more copies of each file you retain, the greater the storage consumption, so you need to determine the data protection/space consumption balance that makes sense for your file system. 118 Chapter 9: Managing cluster resources

129 Managing file systems (Continued) Space HWM Property File system space usage high water mark percentage. A high water mark percentage set at the file system level overrides the cluster-level property File System Space HWM that is set by default for new file systems added to the cluster. When the remaining allocated file system space reaches the high water mark, an alert is raised until the file system space consumption falls below the low water mark percentage. Specify a value from 1 to 100 that is greater than the space-lwm. Keep in mind that any value specified for Space HWM overrides the value of the cluster level File System Space HWM property. Space LWM File system space usage low water mark percentage. A low water mark percentage set at the file system level overrides the cluster-level property File System Space LWM that is set by default for new file systems added to the cluster. Existing file system space alerts are automatically cleared once the low water mark setting is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. Specify a value from 1 to 100. Keep in mind that any value specified for Space LWM overrides the value of the cluster level File System Space LWM property. encryption-algorithm AES128 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 128-bit keys. AES256 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 256-bit keys. This encryption algorithm offers stronger cryptographic protection than AES128. none - no encryption. Chapter 9: Managing cluster resources 119

130 Managing file systems (Continued) Quota (required) Property An administrator-defined maximum file system size limit. HSP monitors and enforces file system quotas and issues alerts when you are approaching quota limits. The quota of a file system is a virtual limitation. Space is allocated only as files are written and stored, so you can create file systems with a much larger quota than the currently available capacity. Specify a value and unit from 128 GiB to 6 PiB. Record Access Time Whether or not the file system should track the last time a file was accessed. Because there is a performance cost associated with tracking access time on frequently changing and accessed files, Record Access Time is set to No. of the file system. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 3. Click OK to save your changes, or Cancel to return to the File Systems page without making changes. 120 Chapter 9: Managing cluster resources

131 Managing file systems Enabling or disabling a file system A file system is enabled by default, meaning that as soon as file systems are created and properly shared out (by adding a share and at least one access rule), clients will have access to the file system by default. Disabling a file system disables all shares associated with the file system, which means that you are disrupting client access to that file system make sure this is what you intend. Shares added to a disabled file system will show as disabled until the file system is again enabled. Some administrators like to create all their file systems at once, but enable and bring them online only as needed. Note: You must disable a file system before you can initiate an fsck on that file system (see "Enabling or disabling a file system" on page 1 for details about disabling a file system). 1. On the File Systems page, select the file system that you want to enable or disable. 2. Click. Chapter 9: Managing cluster resources 121

132 Managing file systems Editing the properties of a file system You can edit the properties of an existing file system, including the quota limit. Quota changes can be made at any time without any client interruption in service. 1. On the File Systems page, click. 2. Select the file system that you want to edit. 3. Enter the property values for the file system that you want to change: Property Name Name of the file system. The name of the file system you provide is also the mount point for the file system, so follow standard UNIX file system naming conventions and rules. File system names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk), / (slash), and ; (semicolon). ID Enabled File system's universally unique identifier (UUID). Whether or not the file system is enabled. By default, the file system is enabled. Specify either Yes or No. 122 Chapter 9: Managing cluster resources

133 Managing file systems (Continued) Property Data Protection Data protection policy set for a file system. By default, the data protection policy is set to raid1_ 3instance_large_io. The current version of the software supports the following values: raid1_3instance_large_io Mirrored with 3 copies or instances of the data, tuned for supporting large streaming writes. This is the ideal for most Big Data Analytics style workloads or general purpose Data Lake. raid1_3instance_small_io: Mirrored with 3 copies or instances of the data, tuned for supporting small random writes. This is ideal for hosting virtual machine images and the best choice for relational databases. raid0_1instance_small_io: A single instance of the data is kept, data is not highly available. It is tuned for supporting small random writes. This is ideal for scratch space where low storage consumption is desired and workloads may be unpredictable, but losing data due to hardware failures would not be catastrophic. HSP automatically ensures that multiple instances of any given file are stored on multiple nodes based on the data protection policy that you specify for a file system. If a node or disk that contains a particular file fails, HSP automatically finds another instance and forwards the client request to another node for one of the file instances. Keep in mind, however, the more copies of each file you retain, the greater the storage consumption, so you need to determine the data protection/space consumption balance that makes sense for your file system. Chapter 9: Managing cluster resources 123

134 Managing file systems (Continued) Space HWM Property File system space usage high water mark percentage. A high water mark percentage set at the file system level overrides the cluster-level property File System Space HWM that is set by default for new file systems added to the cluster. When the remaining allocated file system space reaches the high water mark, an alert is raised until the file system space consumption falls below the low water mark percentage. Specify a value from 1 to 100 that is greater than the space-lwm. Keep in mind that any value specified for Space HWM overrides the value of the cluster level File System Space HWM property. Space LWM File system space usage low water mark percentage. A low water mark percentage set at the file system level overrides the cluster-level property File System Space LWM that is set by default for new file systems added to the cluster. Existing file system space alerts are automatically cleared once the low water mark setting is achieved. The low water mark is typically 10 to 20 percent less than the high water mark. Specify a value from 1 to 100. Keep in mind that any value specified for Space LWM overrides the value of the cluster level File System Space LWM property. encryption-algorithm AES128 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 128-bit keys. AES256 - is a symmetric key algorithm that encrypts and decrypts data in 128-bit blocks, using 256-bit keys. This encryption algorithm offers stronger cryptographic protection than AES128. none - no encryption. 124 Chapter 9: Managing cluster resources

135 Managing file systems (Continued) Property Quota An administrator-defined maximum file system size limit. HSP monitors and enforces file system quotas and issues alerts when you are approaching quota limits. The quota of a file system is a virtual limitation. Space is allocated only as files are written and stored, so you can create file systems with a much larger quota than the currently available capacity. Specify a value and unit from 128 GiB to 6 PiB. Record Access Time Whether or not the file system should track the last time a file was accessed. Because there is a performance cost associated with tracking access time on frequently changing and accessed files, Record Access Time is set to No. of the file system. Free-flowing text, allowing up to 255 characters except ; (semicolon). Tags User-defined tags that can be used to organize and/or filter the response results. Free-flowing text, allowing up to 255 characters except ; (semicolon). 4. Click OK to save your changes, or Cancel to return to the File Systems page without making changes. Chapter 9: Managing cluster resources 125

136 Managing file systems Checking and repairing a file system You can manually initiate running the fsck (file system check) utility to check the consistency of and repair a file system. We recommend running fsck when alert conditions are reported for file systems in the IMPAIRED or ERROR run state. Important: Existing storage issues can impact file systems and cause impairment. Be sure to address any node and disk problems in your cluster before running fsck. Before you can initiate fsck, you must disable the file system (see "Enabling or disabling a file system" on page 1 for details about disabling a file system). The fsck utility will fail if the file system is enabled or if there is already an fsck running on the file system. Upon completion, the event log for the fsck will report the number of errors found, as well as the location of a log file giving a more detailed description of these errors. The log file is copied to the system_info file system in the following location: /debug_logs/fsck If fsck successfully checks and repairs the file system, the HSP software automatically enables the file system. If the fsck does not successfully repair the file system, the file system remains disabled and requires administrator attention to decide on the next course of action. In addition to running an fsck, you can also abort an fsck using any of the HSP management interfaces. 1. On the File Systems page, select the file system that you want to check and repair. 2. Click. 126 Chapter 9: Managing cluster resources

137 Managing file systems Deleting a file system Deleting a file system will delete all files and all instances of the files in the file system. The delete operation also deletes all shares and access rules associated with the file system. Important: Deleting a file system is a permanent removal and you would only be able to recover the files if you have implemented a backup strategy for your cluster. Even though file system deletion happens in the background, this task can take some time to complete, as well as report the reclaimed space statistics. 1. On the File Systems page, select the file system that you want to delete. 2. Click. 3. Confirm that you want to delete the selected file system, by clicking OK. Chapter 9: Managing cluster resources 127

138 Managing shares Managing shares A client cannot access a file system until you explicitly share out that file system and define at least one access rule that grants permission to the share. The current version of the product supports NFSv3 only. A client cannot access a file system until you explicitly share out that file system and define at least one access rule that grants permission to the share. The current version of the product supports NFSv3 only. The Shares page: Provides access to any alert conditions within the cluster Displays the list of shares that have been added to the cluster Provides single-click access to all of the management operations you can perform on shares: Click For help on this topic, see... "Adding a share" on page 130 "Editing the properties of a share" on page 132 "Editing the properties of a share" on page 132 "Managing access rules" on page 134 "Deleting a share" on page 145 Select a share to view its properties. 128 Chapter 9: Managing cluster resources

139 Managing shares Viewing the properties of a share 1. On the Shares page, select the share for which you want to view properties. 2. To view the properties of a share: Click the name of the share if you are in the tile view. Click the ellipses (...) under Details if you are in the list view. Property Name Run State Name of the share. Run state of the file system share: UP Share is available and fully functional. DOWN Share is currently offline. An administrator might have disabled the share. Enabled Whether or not the share is enabled. By default, the share is enabled. File System Name File System ID Name of the file system to which the share is associated. ID of the file system to which the share is associated. Tenant Name Tenant ID Tags Type Name of the tenant associated with the file system share. ID of the tenant associated with the file system share. of the file system share. User-defined tags that can be used to organize and/or filter the response results. Type of file system share. In the current release, only NFS (NFS version 3) is supported and this is the default share type. Chapter 9: Managing cluster resources 129

140 Managing shares Adding a share Before clients can access files in a file system, that file system share must be shared out, which means an associated share and an access rule is required. After you add a share, you must define at least one access rule that grants permission to that share. By default, the HSP software automatically creates a share with the same name as the file system and an access rule called "DefaultAccessRule" that grants read/write access to the share for all client hosts. See "Adding an access rule" on page 136for details. You can override this automatic creation at the cluster level by setting the File System Auto Access property to No. Tip: For ease of management, we suggest that you name the share the same name as the file system. 1. On the Shares page, click. 2. Enter the property values for the share that you want to add: Property Name (required) Name of the share. Supply either a name for the share or enter "autogenerate." The auto-generate option uses the UUID for the name, so we recommend that you specify a more human readable form for the name. Share names can be from 1 to 255 characters and can contain any UTF-8 characters except * (asterisk), / (slash), and ; (semicolon). Enabled Whether or not the share is enabled. By default, the share is enabled. Specify either Yes or No. 130 Chapter 9: Managing cluster resources