Cloudistics On-premises Cloud Platform

Transcription

1 Cloudistics On-premises Cloud Platform User Guide V3.5 Copyright Cloudistics, Inc. All rights reserved. Last updated: March 2018

2 Contents Getting started Welcome to Cloudistics Getting help Using Ignite Accessing Ignite Logging in Supported web browsers for accessing Ignite Working with compute nodes What is a compute node? Understanding compute categories and tags Compute categories Compute tags Viewing and changing compute node details Understanding compute node states Power states Logical states Viewing compute node metrics CPU and memory charts Compute node activity charts Managing compute nodes Connecting to a node console Restarting a compute node Forcing a restart Powering a node on Powering a node off Forcing a node to power off Turning maintenance mode on Page 2

3 Turning maintenance mode off Renaming a node Changing the description of a node Resetting the manager password for the node Adding compute tags to a node Removing compute tags from a node Changing a compute category for a node Moving a node to a different migration zone Customizing the time on a compute node Using a remote NTP Server Prerequisites Editing the chrony.conf file to use a remote NTP server Setting the time zone Configuring the compute node to use NTP Changing the time zone to use default Internet time servers Understanding instance distribution to compute nodes Working with storage pools What are storage blocks and storage pools? Viewing and changing storage pool details Understanding storage pool states Power states Logical states Viewing storage metrics Total storage and storage allocation charts Storage Usage Distribution Storage Overview Storage pool activity charts Managing storage Managing storage controllers Page 3

4 Connecting to the controller console Restarting the storage controller service Managing storage blocks Powering on Powering off Restarting a storage block Renaming a storage block Resetting the manager password Removing a storage block from the storage pool Managing storage pools Renaming a storage pool Adding storage blocks Connecting a storage pool to a migration zone Disconnecting a storage pool from a migration zone Deleting a storage pool Working with migration zones What is a migration zone? Viewing migration zone details Viewing migration zone metrics CPU overview and Memory overview charts Migration zone activity charts Managing migration zones Renaming a migration zone Adding compute nodes Deleting a migration zone Connecting a migration zone to a storage pool Disconnecting a migration zone from a storage pool Working with virtual datacenters What is a virtual datacenter? Page 4

5 Viewing virtual datacenter metrics CPU, memory and storage charts Virtual datacenter activity charts Viewing virtual datacenter allocations Managing virtual datacenters Renaming a virtual datacenter Changing allocations Deleting a virtual datacenter Working with networks What is a VLAN? What is a virtual network? What is a network virtualization (NFV) instance? Firewall settings Set up networking Define VLAN settings Creating a VNET Prerequisites Creating a virtual network in Ignite Configuring a firewall for a VNET (optional) Configuring your edge router Using NetScaler as an NFV Prerequisites Setting up routing to use a Netscaler instance as an NFV Using virtual networks in multi-location organizations Configuring virtual networking for the external routing service Configuring the SDN Router Configuring the vnet1 NFV instance Viewing network details Managing networks Renaming a virtual network Page 5

6 Changing virtual network properties Changing firewall settings Editing the virtual network DHCP service Editing the virtual network routing service Editing the virtual network firewall profile Deleting a virtual network Managing NFV instances Starting the NFV instance Pausing the NFV instance Restarting the NFV instance Forcing restart for the NFV instance Shutting down the NFV instance Forcing shutdown of the NFV instance Unlinking the NFV instance Removing the NFV Deploying the NFV Understanding firewall profiles and overrides Example: Allowing HTTP and HTTPS only Example: Allowing SSH access to a specified range only Configuring a network for CentOS 6 instances Adding multiple network interfaces to an instance Configuring multi-vnic for CentOS instances Configuring multi-vnic for Ubuntu instances Working with application instances What is an application instance? Supported operating systems for applications Viewing and editing application instance details Understanding application instance states Viewing application instance metrics Page 6

7 CPU, memory and storage charts Application instance activity charts Creating and deploying applications Create an instance from a template Create an instance by importing a VM Managing application instances Connecting to the console Pausing an application instance Resuming an application instance Starting an application instance Restarting an application instance Forcing a restart Shutting down an application instance Forcing a shutdown Renaming an application instance Editing an application instance Deleting an application instance Adding a vnic Adding a vdisk Resizing a vdisk After resizing a disk for Windows After resizing a disk for Linux Taking a snapshot Creating a template Inserting guest tools and switching to compatibility/enhanced mode Understanding guest agents and instance modes Inserting guest agent tools Installing or updating a guest agent tools - Windows Installing the drivers Page 7

8 Installing the guest agent tools Installing or updating a guest agent Linux Creating an application instance from ISO Cloning and attaching a vdisk Why cloning and attaching a vdisk is useful Understanding the clone operation benefits How cloning and attaching a vdisk works in Cloudistics Cloning and attaching a vdisk Attaching a vdisk to an instance from its own snapshot, Method Attaching a vdisk to an instance from its own snapshot, Method Attaching a vdisk to an instance from another instance's snapshot Deleting a clone disk Setting a time zone for application instances Configuring NetScaler Prerequisites for configuring NetScaler for the first time Working with application templates What is an application template? What is the Cloudistics application marketplace? What is the template store? Viewing application templates Managing application templates Creating a new instance Editing a template Duplicating a template Deleting a template Allowing access to a share for importing a VM Working with application snapshots and disaster recovery What is an application snapshot? Managing application snapshots Page 8

9 Taking snapshots Reverting a snapshot to an earlier version Start a snapshot as a new instance Renaming a snapshot Viewing vdisks Deleting a snapshot Configuring automatic snapshot settings Deleting all local snapshots Turning off automatic snapshots Invoking Pre- and Post-snapshot scripts Managing Disaster Recovery Requirements Setting up disaster recovery Understanding disaster recovery settings Working with DR snapshots actions Starting a snapshot as a new instance Viewing vdisks Deleting a snapshot Managing DR snapshot settings Editing the DR retention policy Changing the DR location Deleting all DR snapshots Turning off DR Restoring a file from a snapshot Method Method Method Using Rapid Recovery Backing up an Ignite Server 2012 R2 with Rapid Recovery Page 9

10 Performing a Rapid Recovery restore Monitoring and maintaining your environment Monitoring resources and statistics in Ignite Understanding storage activity information (queue depth, IOPS, and latency) Calculating your IO rate Using Centos with FIO Marketplace template to determine latency Managing users Adding users Modifying your user account Configuring notification settings Using Docker in Ignite (example) Editing SDN router network settings Setting up PXE Boot in Ignite Configuring Citrix PVS on Ignite Create PVS Server Setup PXE boot in Ignite (Optional) Setup DHCP outside of Ignite Create instance for PVS desktop Create a Group Policy Object that applies certain settings only to PVS target devices Create PVS target device Create the capture image Prepare the image for production use Create a new PVS target device Forwarding hardware logs to a remote server Understanding CPU load vs CPU in-use metrics Powering down hardware in Ignite Managing encryption and configuring KMIP Understanding encryption in Cloudistics Understanding KMIP compatibility in Cloudistics Page 10

11 Understanding the process for configuring KMIP KMIP server: Vormetric DSM configuration Requirements Creating the Cloudistics Ignite domain and administrator Adding hosts to the domain Configuring storage controller hosts Uploading the client certificate to the host account Testing KMIP communication KMIP server: SafeNet KeySecure configuration Requirements Generating the KMIP server certificate Creating the SSL certificate request Creating a self-signed certificate Generating client certificates Installing the client CA Installing the CA certificate Adding certificates to the trusted CA list Configuring host access Configuring the KMIP Server Managing passwords Revoking host authorization KMIP client configuration Migrating from TPM to KMIP security Active Controller Passive Controller Changing the KMIP-protected password Adding or replacing disks Testing KMIP communication Get the password Page 11

12 Change the password Get the New Password Page 12

13 Getting started Refer to the following topics: Welcome to Cloudistics Using this guide Understanding the process for setting up and using Cloudistics Getting help Welcome to Cloudistics Cloudistics is the first truly composable on premises cloud platform that combines storage, network, compute, and virtualization resources. Cloudistic's integrated and intelligent operating system monitors application performance based on configurable thresholds, giving you unprecedented insight and control over your resources. The Cloudistics platform includes all hardware and software required to deploy and run applications in your on premises cloud and consists of four key infrastructure components: Network Compute Storage Ignite Cloud Controller The tightly integrated network, compute, and storage hardware is provisioned, managed, and controlled through the Ignite Cloud Controller. Once defined and setup, you can use these components together to create and manage your virtual datacenters and applications. Page 13

14 Getting help Whether you have a question about migrating your workloads or provisioning your virtual datacenters on Ignite, we are here to help. Online Chat. Real-time chat is available during business hours, 9am-5pm ET. Initiate a session from our Support page on our website, or more conveniently, directly from the Ignite Cloud Controller management console. Outside of business hours, initiating a support chat will present 2 options: o call our TFN (especially for service-affecting issues), or o fill in a simple form, which will generate an and open a ticket (for non service-affecting issues or general inquiries). Website. User guidance, troubleshooting and knowledge base information can also be found online: Phone. For a suspected or confirmed issue impacting production applications, we recommend calling us at the following: Business hours, Mon-Fri 9am-5pm ET All other times, toll free When calling, please be prepared to provide your name, callback number, company name, site/location, and a general description of the issue. For hardware service, refer to the service description found here. Page 14

15 Using Ignite This guide provides information about using Ignite, the web-based management console, to manage the compute, storage, network and virtualization resources in your environment. Accessing Ignite When your account has been created, you can log in to access the management portal to manage your Cloudistics environment at: Logging in 1. Once your account has been created, you can log in to access the Ignite management console at 2. On the Login page, enter your administrator user account, which is your address. 3. Enter your account password. If you have forgotten your account information, click the link Forgot your Password, and then enter the address that is associated with your account, select the checkbox to confirm that you are not a robot, and click Reset Password. You will receive an with instructions to change your password. Password resets require two-factor authentication. You should also have your mobile phone ready to receive a security code that you will enter to confirm your password reset. 4. Click Login. Supported web browsers for accessing Ignite Ignite works best on the two most recent versions of: Chrome Firefox Microsoft Edge Page 15

16 Note: Older versions of these browsers or browsers not listed above might not be fully compatible. Page 16

17 Working with compute nodes Refer to the following topics: What is a compute node? Understanding compute categories and tags Viewing and changing compute node details Understanding compute node states Viewing compute node metrics Managing compute nodes Customizing the time for a compute node What is a compute node? A compute node is a physical compute resource containing CPU and memory. A compute node occupies one node on a compute chassis. The Ignite management console displays an abstraction of each compute chassis, along with each node within the chassis on the Infrastructure > Hardware page. Understanding compute categories and tags Compute Categories and Compute Tags can be used to specify which application instances can run on a specific compute node. An instance can be migrated from one compute node to another by modifying the instance's compute category or tags. Migrating an instance from one compute node to another involves zero downtime, continuous service availability, and complete transaction integrity. The Ignite VNC console to an instance uses a URL that includes the compute node's IP address; when an instance migrates to a new compute node, the VNC console will automatically close since the URL is no longer valid, and reopening it will open a URL including the new compute node's IP address. Page 17

18 Compute categories Each compute node must belong to exactly one compute category. A compute node's category can be changed only if no application instances are running on it. In the Compute Restraints portion of an application instance's profile, the instance can be configured to run on compute nodes in any category, or only on compute nodes in a specific category. An instance's compute category can be changed while the instance is running. For example, if an instance is running on a compute node in Category1 and the instance's category is changed to Category2, the instance will automatically migrate to a compute node in Category2. In the Ignite management console, the instance will appear to shut down, then start on the other compute node; however, the instance will actually remain running throughout the migration. Compute resources are allocated to virtual datacenters by category. When you create or edit allocations for a virtual datacenter, you enter the amount of CPU and memory to allocate from each category, rather than from individual compute nodes. Compute tags Unlike categories, compute nodes and application instances can have any number of compute tags. An instance can run on any compute node that has at least the same tags as the instance. That is, if an instance has tags Tag1 and Tag2, it can run on any compute node that also has tags Tag1 and Tag2, regardless of any additional tags that compute node has. However, that instance cannot run on a compute node that has Tag1 but not Tag2 or vice versa. Compute tags can be added to or removed from a compute node while instances are running on the node. When removing compute tags, you will be asked to select one of the following action plans for affected instances: Migrate: Ignite will attempt to migrate the instance to another compute node that meets its compute requirements. Relax profile: Ignite will remove the tag requirements from the instance's compute profile. Shut down: Ignite will shut down the instance. Force shut down: Ignite with force shut down the instance. Compute tags can also be added to or removed from an application instance while it is running. If the instance is running on a compute node that lacks the new tag, the instance will automatically migrate to a compute node that has the required tags. In the Ignite Management console, the instance will appear to shut down, then start on the other compute node; but the instance will actually remain running throughout the migration. Page 18

19 Viewing and changing compute node details 1. In the Ignite management console, you can view detailed information about a compute node on the Information tab of a selected node. This includes the compute node's properties, hardware information, and network information. 2. To change a compute node's properties, click the editing icon next to the property you wish to change, or click the Actions button at the top of the screen to view a list of available compute node actions. Note: Hardware information about the compute node is pulled directly from the compute chassis. This information cannot be edited. Network information is based on the compute node's connection to the virtual network router. This information cannot be edited. Understanding compute node states Ignite tracks two types of compute node states. Power states are displayed at the top of each compute node profile page. Logical states are determined based on the compute node's logical relationships to other objects in your cloud infrastructure. Power states Compute nodes can be in one of three power states at all times: Powered on: The compute node is reachable and able to run application instances. Powered off: The compute node has been powered off by the user (for example, for maintenance). Unreachable: The compute node does not respond to the Ignite management console and was not powered off by the user. If a compute node in your cloud infrastructure is unreachable, check your network connectivity or contact customer support. Logical states Unlike power states, logical states may overlap. Logical states include: Unregistered: The compute node has not been registered. Unregistered compute nodes cannot be used to run applications. Cloudistics strongly recommends registering all compute nodes. Registered and Unassigned: The compute node has been registered. It has been assigned to a compute category, but has not been assigned to a migration zone. A compute node must be assigned to a migration zone before it is used to run applications. Registered and Assigned to a Migration Zone: The compute node has been registered. It has been assigned to both a compute category and a migration zone. This is the only logical state in which a compute node can run applications. Page 19

20 Maintenance Mode: The compute node has been registered and assigned to a compute category. Regardless of whether it has been assigned to a migration zone or not, a compute node cannot be used to run applications while it is in maintenance mode. Maintenance mode is a feature that allows you to keep a compute node online, registered, and assigned to a migration zone while simultaneously preventing any application instances from running on the compute node. This is useful when performing certain maintenance activities (for example, changing a compute node's migration zone or compute category). When viewing the compute node's profile page, a yellow Maintenance Mode indicator appears inline with the compute node's compute category and compute tag(s). Viewing compute node metrics CPU and memory charts At the top of the Compute node profile page in the Ignite management console, you can view two donut charts that show the following: CPU - Displays the total physical CPU on the compute node. The %CPU represents the average CPU utilization of the compute node, including hypervisor usage. Memory - Displays Total physical memory in the compute node. Load consists of all data currently held in memory by running and paused instances on the compute node, as well as hypervisor usage. Compute node activity charts On the Activity tab for a selected node, you can view four charts that show recent activity on the compute node. You can adjust the time interval displayed in the charts between the last 6 hours and the last 6 months. Managing compute nodes To perform an action on a specific compute node, open the compute node's profile page, and then click Actions. The actions available on a compute node vary depending on the node's state. Connecting to a node console Select Connect to console from the Actions drop-down menu. Ignite will create a secure VNC connection to the compute node and open the console in a new window. Restarting a compute node 1. Select Restart from the Actions drop-down menu. 1. Select an action plan for each affected application instance. An action plan tells Cloudistics how you want to handle application instances that will be affected by an action you are trying to perform. When you restart a host, you can select one of the following action plans: Page 20

21 Migrate - Ignite will attempt to migrate the instance to another compute node that meets its compute requirements. Relax profile & migrate - Ignite will remove compute category and tag requirements from the instance's compute profile. Ignite will then attempt to migrate the instance to an available compute node in the current virtual datacenter and migration zone. Shut down - Ignite will shut down the instance. Force shut down - Ignite with force shut down of the application instance. 2. Click Continue. When you click Continue, Ignite attempts to perform the selected action plan(s). If an action plan fails, you will be asked to select a different plan. Forcing a restart 1. Select Force restart from the Actions drop-down menu. 2. Click Continue. Note: Ignite will force all running application instances on the compute node to shut down. Powering a node on To power on, select Power on from the Actions drop-down menu. The compute node will immediately power on. Powering a node off 1. To power off, select Power off from the Actions drop-down menu. 2. Select an action plan for each affected application instance. An action plan tells Cloudistics how you want to handle application instances that will be affected by an action you are trying to perform. When you power off a compute node, Ignite offers the following action plans: Migrate - Ignite will attempt to migrate the instance to another compute node that meets its compute requirements. Relax profile & migrate - Ignite will remove compute category and tag requirements from the instance's compute profile. Ignite will then attempt to migrate the instance to an available compute node in the current virtual datacenter and migration zone. Shut down - Ignite will shut down the instance. Force shut down - Ignite with force shut down the instance. 3. Click Continue. Page 21

22 When you click Continue, Ignite will attempt to perform the selected action plans. If an action plan fails, you will be asked to select a different plan. Forcing a node to power off 1. Select Force power off from the Actions drop-down menu. 2. Click Continue. Note: Cloudistics will also force all running application instances on the compute node to shut down. Turning maintenance mode on 1. Select Turn on maintenance mode from the Actions drop-down menu. 2. Select an action plan for each affected application instance. An action plan tells Cloudistics how you want to handle application instances that will be affected by an action you are trying to perform. When you put a compute node in maintenance mode, Ignite offers the following action plans: Migrate: Ignite will attempt to migrate the instance to another compute node that meets its compute requirements. Relax profile & migrate: Ignite will remove compute category and tag requirements from the instance's compute profile. Ignite will then attempt to migrate the instance to an available compute node in the current virtual datacenter and migration zone. Shut down: Ignite will shut down the instance. Force shut down: Ignite with force shut down the instance. 3. Click Continue. When you click Continue, Ignite will attempt to perform the selected action plans. If an action plan fails, you will be asked to select a different plan. Turning maintenance mode off 1. Select Turn off maintenance mode from the Actions drop-down menu. 2. Click Continue. Renaming a node 1. Select Rename from the Actions drop-down menu. 2. Enter a new name for the compute node. Page 22

23 Note: The compute node name must be unique across all compute nodes in your cloud infrastructure. 3. Click Update. Changing the description of a node 1. Select Change description from the Actions drop-down menu. 2. Enter a description for the compute node. 3. Click Update. Resetting the manager password for the node The manager password is used to log into the physical compute node via monitor and keyboard or ssh. 1. Select Reset manager password from the Actions drop-down menu. 2. Enter a new password. 3. Click Reset Password. The updated password will be pushed to the physical compute node. Note: Hardware passwords are not stored on Cloudistics servers. Adding compute tags to a node 1. Select Add compute tags from the Actions drop-down menu. 2. Enter the name of the compute tag(s) you want to add. 3. Click Add Compute Tags. Note: For information about compute tags, see this topic. Removing compute tags from a node 1. Select Remove compute tags from the Actions drop-down menu. 2. Select the compute tag(s) you wish to remove from the compute node. 3. Select an action plan for each affected application instance. An action plan tells Cloudistics how you want to handle application instances that will be affected by an action you are trying to perform. When you remove compute tags from a compute node, Ignite offers the following action plans: Migrate: Ignite will attempt to migrate the instance to another compute node that meets its compute requirements. Page 23

24 Relax profile & migrate: Ignite will remove compute category and tag requirements from the instance's compute profile. Ignite will then attempt to migrate the instance to an available compute node in the current virtual datacenter and migration zone. Shut down: Ignite will shut down the instance. Force shut down: Ignite with force shut down the instance. 4. Click Continue. When you click Continue, Ignite will attempt to perform the selected action plans. If an action plan fails, you will be asked to select a different plan. Changing a compute category for a node To successfully remove a compute node from its original compute category, your infrastructure must have sufficient unused resources in that category. Note: You cannot change a compute category for a compute node while application instances are running on the compute node. Try putting the host in maintenance mode instead. Note: For information about compute tags, see this topic. 1. Put the compute node in maintenance mode. 2. Select Change compute category from the Actions drop-down menu. 3. Select a new compute category for the compute node. 4. Indicate how Cloudistics should reduce the original compute category's allocations (left column). 5. Indicate how Cloudistics should distribute the updated compute category's allocations (right column). 6. Click Change Compute Category. Moving a node to a different migration zone To successfully move a compute node from its original migration zone, your infrastructure must have sufficient unused resources in that migration zone. Note: You cannot change a migration zone for a compute node while application instances are running on the compute node. Try putting the node in maintenance mode instead. 1. Put the compute node in maintenance mode. 2. Select Move to from the Actions drop-down menu. 3. Select a new migration zone for the compute node. 4. Indicate how Cloudistics should reduce the original migration zone's allocations (left column). 5. Indicate how Cloudistics should increase the updated migration zone's allocations (right column). 6. Click Move. Page 24

25 Customizing the time on a compute node Using a remote NTP Server Follow these steps to configure a compute node to use a remote NTP server as the time source. Prerequisites 1. SSH into the compute node with username manager and the password that you set when registering the compute node. 2. Run sudo su and retype the manager password to gain root access. 3. Make a copy of the chrony.conf file before editing it: cp /etc/chrony.conf /home/manager/chrony.conf.original Editing the chrony.conf file to use a remote NTP server 1. Edit the chrony.conf file to use a remote NTP server: vi /etc/chrony.conf 2. Then type i to edit the file. 3. Under # Allow NTP client access from local network, enter: allow where is the NTP server with which you want to sync. Alternatively, you can use the FQDN of the NTP server, and then add an entry for the IP and FQDN of the NTP server in /etc/hosts. 4. Comment out the lines under # Listen for commands only on localhost. by adding # before the line as follows: #bindcmdaddress #bindcmdaddress ::1 5. Under # Use public servers from the pool.ntp.org project. / # Please consider joining the pool ( comment out all the servers listed by adding # before the line as follows: #server 0.centos.pool.ntp.org iburst #server 1.centos.pool.ntp.org iburst #server 2.centos.pool.ntp.org iburst #server 3.centos.pool.ntp.org iburst 6. Add a line: server iburst where is the NTP server you wish to sync with. Alternatively, you can use the FQDN of the NTP server, and add an entry for the IP and FQDN of the NTP server in /etc/hosts. 7. Exit chrony.conf by typing: ESC :wq Setting the time zone 1. To list time zone options, run: timedatectl list-timezones 2. Set the time zone as needed: timedatectl set-timezone region/timezone where region/time zone is the desired time zone (for example, America/New_York) Page 25

26 For example, real time is 9:00am EDT (8:00am CDT). Time on the NTP server is 8:10am CDT, and the required compute node time zone is EDT. After setting the time zone on the compute node to America/New_York and completing the other steps in this article, the compute node's time will be 9:10am EDT. Configuring the compute node to use NTP Configure the compute node to use NTP, and restart services so the changes are implemented: timedatectl set-ntp yes systemctl restart chronyd systemctl restart systemd-timedated.service Changing the time zone to use default Internet time servers These steps set the time zone using the default Internet time servers instead of a remote NTP server. 1. To list time zone options, run: timedatectl list-timezones 2. Set the time zone as needed: timedatectl set-timezone region/timezone where region/time zone is the desired time zone (for example, America/New_York) Understanding instance distribution to compute nodes CPU and RAM allocation is currently used to determine how instances are distributed to compute nodes. The underlying algorithm used by the engine is shown below. The output of the algorithm is the compute allocation factor, which indicates how available a compute node is to run the instance; a compute node with the lowest allocation factor will be selected to run the instance. Page 26

27 Note: Instances can be manually placed on a particular compute node by using compute categories and tags. For information about categories and tags, see this topic. Page 27

28 Working with storage pools Refer to these topics: What are storage blocks and storage pools? Viewing and changing storage pool details Understanding storage pool states Viewing storage pool metrics Managing storage pools What are storage blocks and storage pools? A storage block is a physical storage resource that can be used by application instances. The storage block hardware consists of a set of two storage controllers and one or two storage sleds, all housed in a storage chassis. The Ignite management console displays a representation of each storage block, including controllers and sleds, on the Infrastructure > Hardware page. Figure 1 Storage hardware represented in Ignite Page 28

29 In Ignite, a storage block must be assigned to a storage pool for it to store application instances. In Ignite, the storage pool is where you allocate storage capacity for an application. The storage pool also contains a single template store that can be used by all virtual datacenters connected to the storage pool. You can monitor template store usage on the details page of a storage pool in the Storage Usage Distribution chart. Viewing and changing storage pool details 1. In the Ignite management console, you can view the Storage Pool Information tab, which contains detailed information about each storage pool. This includes the storage pool's properties, hardware information, and network information. 2. To change a storage pool's properties, click the editing icon next to the property you wish to change, or click the Actions button at the top of the portal to view a list of available storage pool actions. Note: Hardware information about the storage pool is pulled directly from the hardware. This information cannot be edited. Network information is based on the storage pool's connection to the virtual network router. This information cannot be edited. Understanding storage pool states The Ignite management console tracks two types of storage pool states. Power states are displayed at the top of each Storage Pool Overview page. Logical states are determined based on the storage pool's logical relationships to other objects in your Cloudistics infrastructure. Power states Cloudistics considers storage pools to be in one of three power states at all times: Powered on: The storage pool is reachable and able to run application instances. Powered off: The storage pool has been powered off by the user (for example, for maintenance). Unreachable: The storage pool does not respond to Ignite and was not powered off by the user. If a storage pool in your cloud infrastructure is unreachable, check your network connectivity or contact customer support. Logical states Unlike power states, logical states may overlap. Logical states include: Unregistered: The storage pool has not been registered. Unregistered storage pools cannot be used by applications. Cloudistics strongly recommends registering all storage pools. Registered and Not Connected: The storage pool has been registered, but has not been connected to a migration zone. Unconnected storage pools cannot be used by applications. Page 29

30 Registered and Connected to a Migration Zone: The storage pool has been registered and is connected to at least one migration zone. This is the only logical state in which a storage pool can be used by applications. Viewing storage metrics Total storage and storage allocation charts At the top of the Storage Pool Overview page in the Ignite management console, you can view storage donut charts that display storage usage. Storage Usage Distribution The Storage Usage Distribution chart breaks down the Total Used storage by the type of data that is using the storage. Template Usage - the amount of space used by templates in the local template store. The space used by a template is roughly equal to the used space on the disks of the template. When an instance is first created on the storage pool, Ignite will save the base image of the instance as a thick provisioned set of blocks. For example, the instance was created from a template; the template is stored as thin provisioned disks, so a thick provisioned base image will be extrapolated from the template. The instance will be created with 0 size and pointed to the base image. If a new instance is created from the same template on the same storage pool, the new instance will point to the base image created for the first instance. The Base Images metric shows how much space is used by these base images. Instances - shows the amount of space used by instances, not including the base image. Local Snapshots - shows the amount of space used by snapshots of instances on this storage pool. DR Snapshots - shows the amount of space used by snapshots taken on other storage pools and DR'ed to this storage pool. Storage Overview The Storage Overview chart displays the Total Storage physically present in the storage pool and how it is allocated. Page 30

31 Total - how much storage is physically present in the storage pool. Allocated - how much storage is allocated to virtual datacenters. Due to oversubscription, this value can be higher than the total storage. Allocated % - what percentage of the total storage is allocated to virtual datacenters. Due to oversubscription, this value can be higher than 100%. Total Used - how much storage is used. Data Reduction - the storage space that is saved by reusing base images for multiple instances. where BaseImage is the size of the base image and Instances is the number of instances using that base image. Storage pool activity charts On the Storage Pool Activity tab, you can view several charts that show recent activity in the storage pool. You can adjust the time interval displayed in the charts between the last hour and the last year. Managing storage Actions can be performed on a storage pool, storage block, or storage controller. The actions available vary depending on the storage pool s state. Managing storage controllers To perform an action on a specific storage controller, either open the Hardware page in the Ignite management console and click beside the storage controller, or click the storage controller and click Actions. Connecting to the controller console Select Connect to console from the Actions drop-down menu. Cloudistics will create a secure VNC connection to the selected controller and open the console in a new window. Page 31

32 Restarting the storage controller service Select Restart storage controller service from the Actions drop-down menu. The Cloudistics storage controller service will be restarted. While it is restarting, instances might become unresponsive. Managing storage blocks To perform an action on a specific storage block, open the Hardware page in the Ignite management console and click "..." beside the storage block. Powering on Select Power on from the Actions drop-down menu. The storage block will immediately power on. Powering off 1. Select Power off on the Actions drop-down menu. 2. Review the list of application instances that will be affected when the storage block is powered off. Select shut down or force shut down for each instance as appropriate. 3. Click Continue. Restarting a storage block 1. Select Restart on the Actions drop-down menu. 2. Review the list of application instances that will be affected when the storage block is powered off. Select shut down or force shut down for each instance as appropriate. 3. Click Continue. Renaming a storage block 1. Select Rename on the Actions drop-down menu. 2. Enter a new name for the storage block. Note: The new name must be unique across all storage blocks in your Cloudistics environment. 3. Click Update. Resetting the manager password The manager password is used to log into the storage controllers via VNC console or SSH. 1. Select Reset manager password on the Actions drop-down menu. 2. Enter a new password. 3. Click Reset Password. Note: The updated password is pushed to the physical storage. Hardware passwords are not stored in Cloudistics. Page 32

33 Removing a storage block from the storage pool 1. Select Remove from the storage pool from the drop-down menu. Note: A storage block can only be removed from a storage pool if all application instances using the storage pool are deleted. 2. Click Remove from the storage pool. Managing storage pools To perform an action on a specific storage pool, click the storage pool in the left-hand column of the Ignite management console to open the Storage Pool Overview page, and then click Actions. Renaming a storage pool 1. Select Rename on the Actions drop-down menu. 2. Enter a new name for the storage pool. Note: The new name must be unique across all storage pools in your Cloudistics environment. 3. Click Update. Adding storage blocks 1. Select Add storage blocks from the Actions drop-down menu. 2. Select the storage block you want to add to the storage pool. A storage block can only belong to one storage pool. 3. Click Add storage block. Connecting a storage pool to a migration zone 1. Select Connect to migration zone on the Actions drop-down menu. 2. Select the migration zone you want to connect to the storage pool. 3. Click Connect Migration Zone. Disconnecting a storage pool from a migration zone 1. On the Storage Pool Overview page, click the Resources & Connectivity tab. 2. Click Disconnect beside the migration zone you want to disconnect from the storage pool. 3. Review the list of application instances that will be affected by the disconnection. Select shut down or force shut down for each instance. These instances cannot be started again until the storage pool is re-connected to the migration zone, or the instances are re-configured to run in a migration zone to which the storage pool is connected. 4. Click Continue. Page 33

34 Deleting a storage pool 1. Select Delete storage pool from the Actions drop-down menu. Note: A storage pool can only be deleted if all application instances using the storage pool are deleted 2. Click Delete Storage Pool. Page 34

35 Working with migration zones Refer to the following topics: What is a migration zone? Viewing migration zone details Viewing migration zone metrics Managing migration zones What is a migration zone? A Migration Zone acts as a boundary around a set of compute nodes. It provides a way to organize a collection of resources virtually and consists of a defined set of compute nodes (with any categories/tags) and connectivity to storage (one or more storage pools). One compute node can belong to one migration zone, and migration zones can connect to one or more storage nodes. Application instances can migrate among nodes within a migration zone but cannot migrate to nodes outside of a migration zone. When an instance migrates from one compute node to another within a migration zone there is no downtime with continuous service availability and complete transaction integrity. When an instance starts in a particular migration zone, it will only utilize the compute nodes in that zone. For a migration zone to support application instances, it must be connected to at least one storage pool. The resources in a migration zone must also be allocated to at least one virtual datacenter before they can be used to run application instances. Viewing migration zone details In the Ignite management console, you can view the Migration Zone Overview page, which displays information about the migration zone, including: Page 35

36 Current software version running on all nodes in the migration zone Every compute category available in the migration zone Every compute tag available in the migration zone Every storage pool connected to the migration zone You can also view the following tabs. Allocations - Each migration zone is an aggregate of resources allocated from compute categories. This tab displays the migration zone's allocations and usage per compute category. It also displays a list of all virtual datacenters to which compute resources have been allocated. Resources & Connectivity - Displays a list of all nodes assigned to the migration zone. The tab also contains a list of all connected storage pools, including usage statistics for each storage pool. Viewing migration zone metrics CPU overview and Memory overview charts In the Ignite management console, you can view two donut charts at the top of the Migration Zone Overview page that show the following: CPU Overview - Displays the Total CPU in the migration zone (that is, the sum of CPU provided by all nodes assigned to the migration zone). This total is broken down into Unallocated CPU and CPU that has been Allocated to virtual datacenters. The Load represents the average CPU utilization across all nodes in the migration zone, including hypervisor usage. This utilization is calculated as a percentage of the total allocated CPU. At the bottom of the donut chart, Ignite displays the migration zone's Oversubscriptions a percentage of the Total CPU. A value over 100% indicates that the migration zone is oversubscribed. Memory Overview - Displays the Total Memory in the migration zone (that is, the sum of all memory provided by all nodes assigned to the migration zone). This total is broken down into Unallocated memory and memory that has been Allocated to virtual datacenters. Allocated memory is further divided into memory In Use by application instances and Free memory; free memory is memory that has been allocated to virtual datacenters, but is not currently in use by applications. Migration zone activity charts On the Migration Zone Activity tab, you can view two charts that show recent activity in the migration zone. You can adjust the time interval displayed in the charts between the last 3 minutes and the last 90 days. Page 36

37 Managing migration zones To perform an action on a specific migration zone, open the Migration Zone Overview page, and then click Actions. Renaming a migration zone 1. Select Rename on the Actions drop-down menu. 2. Enter a new name for the migration zone. Note: The name must be unique across all migration zones in your Cloudistics environment. 3. Click Update. Adding compute nodes 1. Select Add nodes on the Actions drop-down menu. 2. Select the nodes that you want to add to the migration zone. 3. Click Add Nodes. Note: Because a node can only belong to one migration zone, Cloudistics only displays a list of unassigned nodes. If you want to add a node that is currently assigned to another migration zone, you can move the node to another migration zone. Deleting a migration zone CAUTION: This action cannot be undone. 1. Select Delete migration zone from the Actions drop-down menu for the selected migration zone. 2. Click Delete migration zone. Connecting a migration zone to a storage pool 1. In the Ignite management console, go to Infrastructure > Mappings. 2. Next to the storage pool to which you want to connect, on the drop-down menu, click Connect to migration zone. 3. Select the migration zone to which you want to connect. Disconnecting a migration zone from a storage pool 1. On the Migration Zone Overview page, click the Resources & Connectivity tab. 2. For the storage pool from which you want to disconnect, click Disconnect. Page 37

38 3. Review the application instances that will be affected when the flash pool is disconnected. When you disconnect the storage pool from the selected migration zone, all application instances in the migration zone that use this storage pool will be shut down. 4. In the resulting dialog box, click Continue. Page 38

39 Working with virtual datacenters Refer to the following topics: What is a virtual datacenter? Viewing virtual datacenter metrics Viewing virtual datacenter allocations Managing virtual datacenters What is a virtual datacenter? A virtual datacenter is a logical grouping of compute and storage resources allocated from one or more migration zones and one or more storage pools. Applications are deployed in virtual datacenters. A virtual datacenter must have both compute and storage allocations before it can be used to deploy application instances. Viewing virtual datacenter metrics CPU, memory and storage charts At the top of the Virtual Datacenter Overview page in the Ignite management console, you can view three donut charts that show the following: CPU - Displays the CPU Allocated to the virtual datacenter from migration zones. This is broken down into CPU in Use by application instances and Free CPU. Memory - Displays the memory Allocated to the virtual datacenter from migration zones. This is broken down into Memory in Use by application instances and Free memory. Storage - Displays the storage Allocated to the virtual datacenter from storage pools. This is broken down into storage Used by application instances and Free storage. This chart also shows the amount of storage Provisioned to application instances from the virtual datacenter. If the Provisioned storage is greater than the Allocated storage, the storage pool is overprovisioned. Page 39

40 Virtual datacenter activity charts On the Virtual Datacenter Activity tab, you can view several charts that show recent activity in the virtual datacenter. You can adjust the time interval displayed in the charts between the last hour and the last year. If you would like to understand more about storage activity charts, see this topic. Viewing virtual datacenter allocations On the Overview page of a virtual datacenter in the Ignite management console, the Virtual Datacenter Allocations tab contains information about the resources that have been allocated to the datacenter from migration zones and storage pools. The tab includes two sections: Allocations chart - Displays a list of all compute and storage allocations to the virtual datacenter. Compute resources are listed first by migration zone, then by compute category within each migration zone. Storage resources are listed by storage pool. Usage chart - Displays more detailed information about how compute resources in the virtual datacenter are being used by application instances within the datacenter. Usage details are listed first by migration zone, then by compute category. Note: In each chart, compute category has a slightly different meaning: In the Allocations chart, compute category refers to the category to which each host in the migration zone is assigned. In the Usage chart, compute category refers to the compute category type selected by each application in the virtual datacenter. Thus, if application instances in the virtual datacenter do not require a specific compute category, the Usage chart may show statistics for Any Category. Managing virtual datacenters To perform an action on a specific virtual datacenter, open the Virtual Datacenter Overview page, and then click Actions. Renaming a virtual datacenter 1. Select Rename on the Actions drop-down menu. 2. Enter a new name for the virtual datacenter. Note: This name must be unique across all virtual datacenters in your Cloudistics environment. 3. Click Update. Page 40

41 Changing allocations 1. Select Change allocations on the Actions drop-down menu. The Change Datacenter Allocations page opens with your available resources displayed at the top of the page. 2. In the Datacenter Allocation section, adjust the allocations for your virtual datacenter. 3. Click Save Allocation Changes. Deleting a virtual datacenter CAUTION: This action cannot be undone. 1. Select Delete virtual datacenter on the Actions drop-down menu. 2. Click Delete Virtual Datacenter. Page 41

42 Working with networks Refer to the following topics: What is a VLAN? What is a virtual network? What is a network virtualization (NFV) instance? Setting up networking Defining VLAN settings Creating a virtual network Using a virtual network in multi-location organizations Using NetScaler as an NFV Viewing virtual network details Managing virtual networks Understanding firewall profiles and overrides Configuring a network for CentOS 6 instances Adding multiple network interfaces to an instance Using virtual networks in multi-location organizations What is a VLAN? Cloudistics allows you to take advantage of your existing virtual LANs (VLANs) or create new VLANs for use in the Cloudistics Ignite stack. Cloudistics offers site-specific bridged networks based on VLAN tagging for use by application instances. Infrastructure administrators can create VLAN-tagged bridged networks, each tied to a specific site. When an application instance uses these VLAN networks, traffic from those instances is tagged with the VLAN ID. In Ignite, a VLAN network has the following properties: Name, Site, VLAN tag/id. You can use existing VLANs or create new VLANs for use in Ignite. Page 42

43 What is a virtual network? A virtual network (VNET) allows application instances to migrate seamlessly among compute nodes without changing their networking configurations. VNETs can also stretch across multiple migration zones. In such situations, traffic between instances running in different migration zones is encapsulated, which enables you to start instances in a disaster recovery (DR) location without changing IP settings What is a network virtualization (NFV) instance? When you create a VNET, the Ignite management console manages all DHCP and NAT services via an autodeployed application instance called the network function virtualization (NFV) or NFV instance. There is one NFV instance per VNET. This NFV instance acts like any application instance in your cloud infrastructure. It resides in a virtual datacenter and consumes migration zone and storage pool resources. It uses the gateway IP address for your VNET. You can also choose your own application instance to act as a VNET NFV instance. This allows load balancer and firewall applications, such as Netscaler, to run in Cloudistics and act as the VNET gateway. Firewall settings A firewall profile is a set of firewall rules that apply to every application instance running within a virtual network using the profile. These rules can be overridden by application firewall overrides. You can skip firewall settings for now and create a firewall profile later by clicking Firewalls in Ignite. Set up networking Cloudistics allows you to take advantage of your existing virtual LANs (VLANs) or create new VLANs for use in the Cloudistics Ignite stack. You also have the option to set up virtual networks (VNETs) for deploying applications. Each vnic on an application instance can be configured to use any VLAN or virtual network that has access to the virtual datacenter where the instance is deployed. To set up networking, navigate to Networking > Networks from the left side menu. Click New Network and select either New VLAN or New VNET as appropriate. See Figure 2. Page 43

44 Acme.org Cloudistics Dashboard INFRASTRUCTURE Hardware Networks New Network Mappings Allocations Nodes NETWORKING No networks Create networks to micro-segment application instances traffic. Note: At least one VLAN must be created before a VNET can be deployed. Networks Firewalls APPLICATIONS + Figure 2. New Network option on the networks page. Note: VLAN support is a new capability in Cloudistics as of software version 3.5. To take advantage of this feature, existing customers with earlier versions must go through an on-site remanufacturing process for their network routers (which involves a restart of the Ignite infrastructure). After the update, all existing vnics using bridged mode will be reset, and, either a VNET or VLAN, must be selected to start those instances. Contact Cloudistics Support for more information. Define VLAN settings Cloudistics offers site-specific bridged networks based on VLAN tagging for use by application instances. Infrastructure administrators can create VLAN-tagged bridged networks, each tied to a specific site. When an application instance uses these VLAN networks, traffic from those instances is tagged with the VLAN ID. A VLAN network has the following properties: Name, Site, VLAN tag/id. You can use existing VLANs or create new VLANs for use in Ignite. When you click New VLAN on the Networks page, you will need to define VLAN properties. See Figure Figure 3 Define VLAN properties when adding a new VLAN Page 44

45 1. Enter a name for the VLAN. 2. Select the site where the VLAN should be available. 3. Enter a VLAN Tag (the VLAN ID). If want to use an existing VLAN, enter the VLAN ID of your existing VLAN network. Note: Each site can use a unique set of VLAN tags. Note: If this VLAN must be available at other sites, you should create a new VLAN in Ignite for each of the sites but with the same VLAN ID. If you have more than one location with different VLAN tags and create a migration zone that spans the different locations, then the application could potentially start on a compute node in a location that does not support the given VLAN tag. However, if the different locations share a VLAN tag, then as long as one of the locations VLANs is chosen, then the migration zone spanning multiple locations will work when starting the application. 4. Click Create VLAN. Note: For application instances to use the VLAN, you would select the VLAN while configuring the application instance vnic(s). Creating a VNET Prerequisites Before you create a virtual network, you should: Determine how large your subnet should be Determine how many CPU cores you will need and how much memory you would like to allocate to your VNET s Network Function Virtualization (NFV) instance. (The NFV instance acts as a DHCP server and gateway for instances in the VNET.) Have created a migration zone. Have created a storage pool. Have created a virtual datacenter. Have created a firewall profile (if needed) Creating a virtual network in Ignite To create a virtual network in Ignite, follow these steps. 1. In Ignite, click Networking > Networks in the side menu. 1. On the Networks page, click New Virtual Network, and then click New VNET. Page 45

46 2. In the Properties section, you will need to enter your VNET name, network address (IP address preceding your usable IP range), netmask, and default gateway. See Figure 4 and complete the following steps. Figure 4. Virtual network properties in the create VNET page 1. Select the type of network service as one of the following. Automatic - A network functional virtualization (NFV) instance for this VNET will be created for you. Note: An auto-deployed NFV instance will act as a DHCP server and gateway for instances in the corresponding VNET. It will be assigned two IP addresses: the internal interface IP address specified as the VNET s gateway, and the outside interface IP address, which will be determined by the Routing Service settings. Manual - You will need to create and configure an NFV instance to provide network services for this VNET. See the section, Deploy VNET - custom NFV instance, for more information. Define VNET settings - auto-deployed If you selected Automatic in the preceding step, you need to define the DHCP settings for this VNET. See Figure 5 and the following steps. Page 46

47 Figure 5. Configuring DHCP service for VNET with auto-deployed NFV instance 1. Specify the part of this VNET s IP address range that should be served by DHCP. This DHCP range must be within the usable IP address range of this VNET. 2. Optionally, you may specify advanced DHCP settings such as lease time. This is how long your DHCP lease is retained on your client. The default is set to one day (86400 seconds), but you can override it. 3. You may also choose a domain name, static binding, and primary and secondary (redundant) DNS servers for DHCP. These are all optional and can be altered later. Cloudistics recommends entering a Primary DNS server and keeping the other default settings. 4. If you plan to assign fixed IP addresses via DHCP, you can use the Static Binding option. Once instances using your VNET have been created, you can enter their MAC addresses and desired IP addresses in the Static Binding section. These IP addresses must be in your DHCP range. The specified instances will request the specified IP addresses from DHCP. To avoid issues with any outstanding DHCP leases, make sure you also set the IP address specified in this section as the static IP address within the instance's guest OS. Next, you should define NFV Instance settings as needed. See Figure 6. Page 47

48 1 2 Figure 6. Location, provisioning, and other settings for auto-deployed NFV instance 1. Select the virtual datacenter, storage pool, and migration zone where the auto-deployed NFV instance should run for this VNET. 2. Specify CPU, memory, and any optional compute constraints for the NFV instance. The Outside Interface (Routing Service) settings define the external network through which all traffic of this VNET will communicate out. You can select the outside interface only after a migration zone has been selected for the NFV. You have the option to choose VLAN or VNET as the external networking mode. Note: If you choose a VLAN network for the external network, a static route needs to be created on your core router. For this external IP address, you can choose between DHCP (IP address will be automatically assigned) or Static (to assign an IP address of your choosing). See Figure Figure 7. Configuring outside interface routing service Page 48

49 1. For the external networking mode, select VLAN (default) or VNET as appropriate and then select the actual network to use for external traffic. 2. Select DHCP or Static as appropriate. If you select Static, enter the IP information. Now you can click Create Virtual Network (see Figure 8). It may take a few minutes to deploy and configure the VNET. Figure 8. Create virtual network option at the bottom of the page Deploy VNET - custom NFV instance 1. If you chose to use your own NFV instance for the VNET, (that is, you create the VNET with the Manual option in the Network Services section of the VNET creation page), you need to create and configure your NFV instance. At least one of the vnics for this instance must use this VNET. 2. After configuring the application instance, click Link NFV Instance. See Figure 9. Figure 9. Link NFV instance option on a VNET with custom NFV instance 3. Start the custom NFV instance to complete the VNET creation. 4. To link a custom NFV instance to a VNET, select the virtual datacenter and the NFV instance that you created, and then click Save. See Figure 10. Page 49

50 Figure 10. Linking a custom NFV instance to the VNET Configuring a firewall for a VNET (optional) In Ignite you can specify a firewall profile for any VNET. An Ignite firewall profile is a set of rules. You would need to create a new firewall profile and associate it to this VNET. See Figure 11 and the following steps. 2 1 Figure 11. Specifying a firewall for the VNET 1. Select a pre-created firewall for your VNET to use. 2. Create a new firewall by selecting Firewall profiles link. Configuring your edge router The Edge Router is required to be configured when VNETs are created. For virtual networking to work successfully in infrastructures without BGP, your edge router must be configured to recognize the IP addresses in each virtual network. Ensure that your edge router has a rule to route all traffic coming to each virtual network s IP address range to the Ignite router s IP address. Page 50

51 Using NetScaler as an NFV Prerequisites You need to complete the following steps before configuring NetScaler as an NFV instance: 1. Create a Vnet and choose "manual" as your Network services. 2. Download/create a Netscaler application with bridged networking for vnic0. Then, add a second vnic, selecting virtual networking and the VNET you just created. 3. Go to the VNET you created and click "link NFV instance". Choose the Netscaler instance you created 4. Follow the steps to configure NetScaler as described here. You are configuring networking for vnic0, so give it an IP in the bridged network and configure other settings accordingly Setting up routing to use a Netscaler instance as an NFV Follow these steps. 1. From your SDN router, enter this command: /usr/share/cloudistics/export_bgp_route.sh "add <Network address of vnet> <mask range of vnet><netscaler vnic0 IP address>" Example: /usr/share/cloudistics/export_bgp_route.sh "add "] Note: The network address is the IP address before your VNET's usable IP range 2. From NetScaler in your browser, enter the IP address you configured for Netscaler into your browser and log in. 3. Under the configuration tab, click subnet IP address. 4. Add the gateway IP address and subnet mask of your VNET, and click Create. 5. Click the Save button. 6. Remove all other IPs under the subnet IP address and click the Save button. You should be able to ping your VNET gateway from outside of the virtual network if your NetScaler NFV is properly set up. Note: NetScaler is not a DHCP server. Within the operating system of an instance using the NetScaler VNET, configure a static IP within the VNET IP range and set the gateway to the VNET's gateway. Page 51

52 Using virtual networks in multi-location organizations The following example demonstrates the use of virtual networks (VNETs) in an organization with multiple locations. In this example: LocationA has SwitchA, ComputeA and StorageA. Migration-ZoneA, virtual-datacentera and Storage-PoolA are only allocated hardware from LocationA LocationB has SwitchB, ComputeB and StorageB. Migration-ZoneB, virtual-datacenterb and Storage-PoolB are only allocated hardware from LocationB. The edge router is configured to send traffic to SwitchA, and traffic to SwitchB. A VNET (vneta) using the range can only work properly if the VNET services application is deployed on hardware associated with SwitchA. That is, when creating the VNET Migration-ZoneA, virtual- DatacenterA and Storage-PoolA should be selected. A VNET (vnetb) using the range can only work properly if the VNET services application is deployed on hardware associated with SwitchB. That is, when creating the VNET Migration-ZoneB, virtual- DatacenterB and Storage-PoolB should be selected. An instance using vneta must be deployed on hardware associated with SwitchA for its networking to function properly. That is, when creating the instance Migration-ZoneA, virtual-datacentera and Storage-PoolA should be selected. An instance using vnetb must be deployed on hardware associated with SwitchB for its networking to function properly. That is, when creating the instance Migration-ZoneB, virtual-datacenterb and Storage-PoolB should be selected. Configuring virtual networking for the external routing service When creating a virtual network (VNET) in Cloudistics version 3.3 and later, additional configuration is required if "Virtual Networking" is chosen for the external Routing Service: a static route must be added to the SDN Router and the Network Function Virtualization (NFV) instance of the VNET acting as a routing service. Note: If you choose a VLAN network for the external network, a static route needs to be created on your core router. In the following example, a user is creating Vnet2 that will use Vnet1 as the Routing Service (external) interface. Page 52

53 Vnet1: Vnet IP Range: /24 External Routing Service IP (bridged): Internal IP: Vnet2: Vnet IP Range: /24 External Routing Service IP (Vnet1): Internal IP: Configuring the SDN Router Follow these steps to add a static route to the SDN router. 1. SSH into the SDN Router with the username manager and the password that was set when the hardware was registered. 2. Run the command, sudo su, and retype your manager password to become root 3. Run the command: /usr/share/cloudistics/export_bgp_route.sh "add <network address of Vnet2> <subnet mask of Vnet2> <external IP of Vnet1>" In this example, the command would be: /usr/share/cloudistics/export_bgp_route.sh "add " Configuring the vnet1 NFV instance Follow these steps to add a static route to the Network Function Virtualization (NFV) instance of the VNET acting as a routing service. 1. Open the console to the NFV instance for Vnet1. 2. Log in as root with password cloudistics. 3. Enter the command: vi /etc/sysconfig/network-scripts/route-internal and type i to edit the file. 4. Enter the command: <vnet IP range of Vnet2> via <external IP of Vnet2> dev internal In this example, the entry would be: /24 via dev internal 5. Press ESC and enter :wq to write changes and close the file. 6. Run ifup internal. Page 53

54 Alternatively, you can use an application instance as a NFV. Learn more here. Viewing network details You can access details about a network by going to Networking > Networks in the Ignite management console, and then selecting the virtual network for which you want to view details. The resulting Overview page contains detailed information about the selected network. Managing networks You can access details about a network by going to Networking > Networks in the Ignite management console, and then selecting the virtual network for which you want to view details. The resulting Overview page contains detailed information about the selected virtual network. This includes information about the virtual network's auto-deployed services (DHCP and NAT), the network services application, and the firewall profile (if any). To perform an action on a virtual network, on the Overview page for a virtual network, click Actions.. Renaming a virtual network 1. Select Edit name on the Actions drop-down menu. 2. Enter a new name for the virtual network. Note: This name must be unique across your Cloudistics infrastructure. 3. Click Update. Changing virtual network properties 1. Select Edit properties on the Actions menu. 2. Change the VNET IP range settings as needed. 3. Click Save. Note: These changes will not take effect on the clients until they renew their DHCP leases. Changing firewall settings 1. Select Edit firewall profile on the Actions menu. 2. Change the settings for the virtual network as needed. 3. Click Save. Page 54

55 Editing the virtual network DHCP service Note: These changes will not take effect on the clients until they renew their DHCP leases. 1. Click Edit DHCP service from the Actions drop-down menu. 2. Change the VNET's DHCP settings as needed. 3. When you are ready, click Save. After adding, changing, or removing a static binding from an instance that already has a DHCP IP, the DHCP server and client will still have the old lease cached, so additional steps are required: 4. Open a console to the NFV instance. 5. Run vi /var/lib/dnsmasq/dnsmasq.leases and type i to edit the file. 6. Remove the line containing the old IP of the instance whose static binding is being changed. Enter ESC and :wq to save the changes. 7. Run systemctl restart dnsmasq 8. Open a console to the instance whose static binding is being changed 9. For Linux instances: run rm /var/lib/networkmanager/dhclient-*-<interfacename>.lease to remove the old DHCP lease file, and nmcli con up <Interface Name> to bring up the interface. 10. For Windows instances: Open Network and Sharing Center and click Change adapter settings. Select the network adapter and click Disable this network device, then click Enable this network device. Editing the virtual network routing service 1. Click Edit routing service from the Actions drop-down menu. 2. Change the VNET's routing service settings as needed. See here for more information if you are using Virtual Networking for the routing service. 3. When you are ready, click Save. Editing the virtual network firewall profile 1. Click Edit Profile from the Actions drop-down menu. 2. Change the VNET's firewall profile as needed. See here for more information on firewalls. 3. When you are ready, click Save. Deleting a virtual network CAUTION: This action cannot be undone. 1. Select Delete virtual network on the Actions drop-down menu. 2. Click Continue to confirm. 3. Click Delete Virtual Network. Page 55

56 Managing NFV instances Once the network is selected and the Overview page is open, on the Actions menu you can select a variety of actions to perform on the NFV instance. Note: Some actions are dependent upon the NFV instance's state. Starting the NFV instance 1. Click NFV Instance > Start on the Actions drop-down menu. 2. Review the location and provisioning information for the network services application. If a migration zone has not already been selected for the instance, select one from the drop-down menu. 3. Click Start. Pausing the NFV instance 1. Click NFV Instance > Pause on the Actions drop-down menu. 2. Click Pause Instance. Restarting the NFV instance Note: Restarting the NFV can cause network connectivity interruptions. 1. Click NFV Instance > Restart on the Actions drop-down menu. 2. Click Restart Instance. Forcing restart for the NFV instance Note: Forcefully restarting NFV can cause network connectivity interruptions. 1. Select NFV Instance > Force restart on the Actions drop-down menu. 2. Click Force Restart Instance. Shutting down the NFV instance Note: Shutting down NFV can cause network connectivity interruptions. 1. Select NFV Instance > Shut down on the Actions drop-down menu. 2. Click Shut Down Instance. Forcing shutdown of the NFV instance Note: Forcefully shutting down NFV can cause network connectivity interruptions. 1. Select NFV Instance > Force shut down from the Actions drop-down menu. Page 56

57 2. Click Force Shutdown of Instance. Unlinking the NFV instance 1. Click Actions > NFV Instance > Unlink NFV instance 2. Click Unlink NFV Instance. Removing the NFV Note: Removing the NFV instance will cause network connectivity interruptions for the VNET until a new NFV instance is deployed. 1. Select Remove NFV Instance from the Actions->NFV Instance drop-down menu. 2. Click Remove NFV Instance. Deploying the NFV Note: After removing the NFV instance, a new one can be deployed by clicking Deploy NFV Instance. 1. Select the Virtual Datacenter, Storage Pool, and Migration Zone in which you want your NFV instance to run. 2. Click Save. Understanding firewall profiles and overrides An Ignite Firewall Profile or Override is a set of Rules. A Profile can be applied to a virtual network, and an override can be applied to an Application. Two rules are available by default in a virtual network firewall profile that: Allow outgoing and incoming UDP traffic to destination ports Allow traffic from instances to the virtual network and DHCP requests from instances to the VNET, respectively. CAUTION: If these rules are deleted and rules are added that block this traffic, the virtual network will not function normally. 1. (For Cloudistics version 3.3 and later) For VNETs with a firewall profile, create an App Firewall Override profile that Allows Incoming ALL from the Source IP Range that matches the Usable IP Range of the VNET. 2. Click the VNET's NFV instance, and go to its Profile tab. Beside the internal vnic (the one whose IP corresponds to the VNET's gateway), click "...", "Edit firewall override", and then select the App Firewall Override created in the previous step. Page 57

58 Application Firewall Override rules are evaluated prior to the virtual network firewall profiles rules, so they override the virtual network firewall rules if there is a conflict. Firewall Rules follows this iptables rule chain: Rules are evaluated in order from top to bottom; If the rule matches, it executes the specified action; If the rule does not match, it moves to the next rule; All communication is allowed unless otherwise specified. Example: Allowing HTTP and HTTPS only Figure 12 Example rule allowing HTTP and HTTPS only Example: Allowing SSH access to a specified range only Figure 13 Example rule allowing SSH access to a specified range only Configuring a network for CentOS 6 instances There is a known issue with cloning CentOS 6 machines wherein the resulting machine retains information regarding the original's eth0 network interface (including MAC address). As a result, it labels the new network interface as eth1 and is unable to start it because there are no network scripts referencing eth1. This can be resolved in one of two ways, described as followed Page 58

59 Method 1: Select Run sysprep when creating a template of a CentOS 6 instance. This will prevent the problem from occurring. Method 2: After creating a CentOS 6 instance without running sysprep, complete the following steps. 1. In the Ignite management console, go to the profile of the problematic instance. Make note of the MAC Address (we will use 76:00:00:00:00:00 for this example). 2. Log in to the CentOS instance and run: vi /etc/udev/rules.d/70-persistent-net.rules There will be two entries: eth0 and eth1. Type i to input data, and delete the entry with the MAC address that does not match the one from the first step above. If the remaining entry is for eth1, change it to eth0. Press ESC and type :wq to write the changes and exit. 3. Run: ls /etc/sysconfig/network-scripts If ifcfg-eth1 is present, delete it as follows: rm -d /etc/sysconfig/network-scripts/ifcfg-eth1 4. Run the command: vi /etc/sysconfig/network-scripts/ifcfg-eth0 5. Type i to input data, and add a HWADDR={MAC Address} entry. Set ONBOOT=yes if it is not set already. Press ESC and type :wq to write the changes and exit. Figure 14 Example modification of /etc/sysconfig/network-scripts/ifcfg-eth0 6. To start the network interface, run the command: ifup eth0 Page 59

60 Adding multiple network interfaces to an instance To add multiple network interfaces (Multi-vNIC) to an instance, follow these steps. 1. Shut down the instance. 2. On the instance's Profile page under Networking, click Add vnic, and select the options for the vnic you wish to add. If it is a CentOS instance, see this topic, Configuring multi-vnic for CentOS instances, for additional instructions. If it is an Ubuntu instance, see this topic, Configuring multi-vnic for Ubuntu instances, for additional instructions. Note: Up to 64 vnics (vnic 0 - vnic 63) can be added to a single instance. Configuring multi-vnic for CentOS instances For CentOS instances to be able to use multiple vnics, the network interfaces must be configured within the operating system. Follow these steps to configure the network interfaces. 1. In the Ignite management console, go to the Profile of the problematic instance. Make note of the MAC Addresses for each vnic (in this example, B4:D1:35:00:03:6A for vnic 0 and B4:D1:35:00:03:6B for vnic 1). 2. Log into the CentOS instance. 3. Run: cd /etc/sysconfig/network-scripts and then: ls If there are no ifcfg-eth# files for all of the vnics, those files will need to be created. For example, if two vnics are attached, there should be ifcfg-eth0 and ifcfg-eth1. 4. To create a copy of the ifcfg-eth0 file, run: cp ifcfg-eth0 ifcfg-eth1 5. Edit the ifcfg-eth# files as follows: Run: vi /etc/sysconfig/network-scripts/ifcfg-eth0 Type i to input data, and add the HWADDR={MAC Address} entry noted in the first step. Set ONBOOT=yes if it is not set already. Press ESC and type :wq to write the changes and exit. Page 60

61 Figure 15 Example modification of /etc/sysconfig/network-scripts/ifcfg-eth0 Run: vi /etc/sysconfig/network-scripts/ifcfg-eth1 Type i to input data, and add the HWADDR={MAC Address} entry noted in the first step. Set ONBOOT=yes if it is not set already. Change DEVICE=eth0 to DEVICE=eth1, and remove the UUID line so that it does not conflict with eth0. Press ESC and type :wq to write the changes and exit. Figure 16 Example modification of /etc/sysconfig/network-scripts/ifcfg-eth1 Run ifup eth0 and ifup eth1 to start the network interfaces. Note: These instructions will bring up the network interfaces and allow them to receive IPs. If additional networking configuration is required, please refer to CentOS documentation. Configuring multi-vnic for Ubuntu instances Follow these steps to configure multiple network interfaces for Ubuntu instances. 1. Log into the Ubuntu instance and become root. Page 61

62 2. Run: ip a Make note of the interface names. 3. Edit /etc/network/interfaces as follows. Repeat the lines for the primary interface, replacing the interface name with that of the secondary interface. For example, to enable interface ens4 add these lines: auto ens4 iface ens4 inet dhcp 4. Run service networking restart. Note: These instructions will bring up the network interfaces and allow them to receive IPs. If additional networking configuration is desired, please refer to the Ubuntu documentation Page 62

63 Working with application instances Refer to the following topics: What is an application instance? Supported operating systems for applications Viewing application instance details Understanding application instance states Viewing application instance metrics Managing application instances Creating an application instance from ISO Understanding guest agents Setting a time zone for application instances What is an application instance? An application instance is one or more virtual machines (VMs) grouped together to realize a specific set of software capabilities for an end user. For example, an application instance might be a simple VM running Windows 8; and another application instance might be a VM running CentOS 7.2 with WordPress. In Ignite, you can create an application instance in the following ways: Download a template from the Cloudistics Application Marketplace, customize the template as needed, and deploy a new instance with just a few clicks. Create a new application instance from an existing template stored in your template store. Import a VM from a.vmdk,.vhd,.vhdx,.img,.ova,.qcow2,.qed,.raw, or.vdi as a template, and then use the template to create a new instance. Install a new instance from an ISO or custom installer. Note: Your disk module needs to be compatibile with Ignite. You might need to configure the kernel module for proper disk support, with SATA or Virtio. Page 63

64 Example: The following command can be used in some cases to add the VirtIO kernel modules on boot: dracut add-drivers virtio_pci virtio_blk virtio_scsi virtio_net -f -v For more information about application templates, the marketplace and the template store, see the topic, Working with application templates. Supported operating systems for applications Ignite is constantly increasing application operating system compatibility. Ignite currently supports applications running the following Operating Systems: Red Hat (CentOS/RHEL/Scientific Linux) 6/7 (64-bit) Ubuntu LTS Desktop/Server (64-bit) Microsoft Windows 7/8/8.1/10 (32-bit/64-bit) Microsoft Windows Server 2008R2/2012/2012R2/2016 (64-bit) FreeNAS 9.10 FreeBSD 10.3 Netscaler Viewing and editing application instance details 1. In the Ignite management console, you can view application instances by going to Applications > Instances. 2. On the Application Instances page, under the Name column, select the instance for which you want to view details. 3. On the application s page, click the Profile tab. 4. You can view detailed information about the application instance, including the general information, compute and storage resource information, networking information, firewall rules, and other settings. 5. To edit the details on this tab, click the Edit icon next to the setting you want to change, or, on the Actions menu, click the Edit Instance and then select the setting you want to edit. For more information about actions you can perform on an instance, see the topic, Managing application instances. Understanding application instance states On the Application Instances page, you can view application instance states in the list of application instances. You can also click to view the details of an application instance, and the state will be displayed at the top of the page. Application instances can be in one of the following states: Page 64

65 Powered on: The instance is operational after a successful start or restart. (Instances do not have to run a guest agent; therefore, an instance can show running before the guest agent starts or the instance is fully booted. Paused: The instance is reachable, but paused. Application instances continue to use memory resources while paused. Shutdown: The instance has been properly shut down by the user. Unreachable: The instance does not respond to Ignite and was not shut down by the user. If an instance in your cloud infrastructure is not responding, check your network connectivity or contact customer support. Unresponsive: The heartbeat message sent by the guest agent is no longer being received by the compute node; for example, if the guest agent is seen as missing or out of date. This can often be resolved by restarting the guest agent service or a clean shutdown and start of the instance. Temporary revert: The instance has been reverted to a snapshot and the reversion has not been made permanent Changing state: The instance is in the process of changing from one state to another Viewing application instance metrics CPU, memory and storage charts On the Application Instances page in the Ignite management console, when you select an application instance to view details, you will see three donut charts at the top of the page that show the following information. CPU - Displays the CPU Provisioned to the instance when it was created. The Load represents the average CPU utilization of the instance. Memory - Displays the Memory Provisioned to the instance when it was created. The footprint represents the data currently held in memory by the instance. Storage - Displays the Storage Provisioned to the instance when it was created, as well as the amount of storage Used. Page 65

66 Application instance activity charts You can click the Activity tab to view several charts showing recent activity for the instance. You can adjust the time interval displayed in the charts between the last 3 minutes and the last 90 days. For more information about what is shown in the storage activity charts, see this topic. Creating and deploying applications Cloudistics offers several ways to create application instances: Create from Template - Use a pre-defined template to quickly deploy new application instances. These instances function as thin-provisioned clones that share a common base image, reducing storage usage across a virtual datacenter. Create from Installer - Use an ISO file to create a new application instance. Import a VM - Import a file with extension VMDK, VHD, VHDx, IMG, OVA, QCOW2, QED, RAW, or VDI as a template and use the template to create a new application instance. Note: Cloudistics ships select templates with every Ignite solution. You can use one of those templates to create a new application instance or download more templates from the Cloudistics Application Marketplace. Create an instance from a template 1. In the left side menu, under Applications, click Templates to view the list of application templates available in your infrastructure. 2. Select a template and click New Instance. See Figure 17. Figure 17. Instantiating a template by selecting New Instance Page 66

67 3. Enter a unique name and optional description for the application instance. Figure 18. Name and description for the new instance 4. Select the virtual datacenter where the application instance will run. Note: The virtual datacenter selection and the mapping rules (defined in Error! Reference source not found.) determine the migration zone and storage pool resources available for the application instance. You can also deploy more than one instance or tag the instance with a group name from this page. See Figure 19. Figure 19. Selecting the VDC, number instances, and group name options while instantiating a template The next step is to define storage, compute and network settings. See Figure 20. Page 67

68 Figure 20. Selecting storage pool, migration zone, and network options during template instantiation 5. Select the storage pool on which this application instance should be stored. 6. Select the CPU, memory provisioning and the migration zone in which this application instance should run. 7. Select the network mode as either VNET or VLAN. If you created a virtual network in the section, Error! Reference source not found., you can select it by using the VNET option. Else, you can select the VLAN option. Next you should select instance settings and click Create Instance. See Figure 21. Page 68

69 Figure 21. Additional instance settings in template instantiation Create an instance by importing a VM In Ignite, you can import an existing VM (in VMDK, VHD, VHDx, IMG, OVA, QCOW2, QED, RAW, or VDI formats) as a template and use that template to create new application instances. Navigate to Applications > Templates and click Import VM on the top right of the page. See Figure 22. Figure 22: "Import VM" option on the Application Templates page Page 69

70 You will need to enter a name and other information for the new application template, which will be created from the imported VM. See Figure 23 and the following steps Figure 23: Specifying details of the template to be created from the imported VM 1. Enter a unique name for the new application template. 2. Click Select Virtual Disk Files. 1. Select the OS type and version of the VM being imported and click Continue. See Figure 24. Page 70

71 Figure 24: Selecting OS type and version of the VM being imported 2. Connect to the external store where the VM files are saved. Select NFS or CIFS and provide the required information to connect to the external store, then click Connect. See Figure 25. Figure 25: Connecting to the external store where the VM files are stored 3. Choose VM files from the files listed at this location. Select the root disk file(s) you wish to import, then click Add Selected. 3. Select whether the VM has virtio drivers installed. If you are unsure, choose VM does not have virtio drivers installed Page 71

72 4. Click Save as Application Template. Once the application template is created, you will be able to view it in your list of templates at Applications > Templates. Managing application instances Actions you can perform on an instance are available in the Actions menu at the top right corner of an application instance overview page in Ignite. 1. On the Application Instances page in Ignite, in the Name column, select the instance you want to manage. 2. Click Actions. The actions available for an instance will vary depending on the state of the instance. Connecting to the console Select Connect to console on the Actions drop-down menu. Page 72

73 Ignite will create a secure VNC connection to the instance and open the console in a new window. For the console to open, the firewall must allow incoming connections from the internal network to the compute node's (Router physical network) through ports TCP , inclusive. Pausing an application instance 1. Select Pause on the Actions drop-down menu. 2. Click Pause Instance. 3. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Pause button at the top Resuming an application instance Select Resume on the Actions drop-down menu. The instance will resume automatically. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Start/Resume button at the top Starting an application instance Select Start on the Actions drop-down menu. The instance will start automatically. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Start/Resume button at the top Restarting an application instance 1. Select Restart on the Actions drop-down menu. 2. Click Restart Instance. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Shut Down button at the top and selecting Restart. Forcing a restart 1. Select Force restart on the Actions drop-down menu. 2. Click Force Restart Instance. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Shut Down button at the top and selecting Force restart. Shutting down an application instance 1. Select Shut down on the Actions drop-down menu. Page 73

74 2. Click Shut Down Instance. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Shut Down button at the top and selecting Shut down. Forcing a shutdown 1. Select Force shutdown from the Actions drop-down menu. 2. Click Force Shut Down Instance. Note: This can also be done as a bulk-action by selecting multiple instances from the Application Instances page, clicking the Shut Down button at the top and selecting Force shutdown. Renaming an application instance 1. On the Actions drop-down menu, hover the mouse over Edit Instance and select Edit name. 2. Enter a new name for the application instance. Note: The new name must be unique across all instances in your Cloudistics infrastructure. 3. Click Update. Editing an application instance On the Actions drop-down menu, hover the mouse over Edit Instance and then select the option you want to edit. Make the change(s) and click Save. Available options to select are: Edit name Edit description Edit CPU provisioning Edit memory provisioning Edit compute constraints Change virtual datacenter. Note: If the virtual datacenter is changed, the migration zone must also be changed. Change migration zone Disable automatic recovery Disable VT-x flag Edit timezone configuration Edit boot order Page 74

75 Deleting an application instance CAUTION: This action cannot be undone. 1. Select Delete from the Actions drop-down menu. 2. Click Delete Instance. Note: You can also delete multiple instances at once by selecting the instances on the Application Instances page, clicking the... button at the top, and then selecting Delete. Adding a vnic 1. Select Add vnic from the Actions drop-down menu. 2. Enter the settings for the vnic. 3. Click Add vnic. Note: You can only add a vnic while an instance is shut down. Adding a vdisk 1. Select Add vdisk from the Actions drop-down menu. 2. To add a new vdisk: a. Select New vdisk. b. Enter a name and size for the vdisk. c. Click Add vdisk. 3. To add a vdisk clone: a. Select Add vdisk clone. b. Select the snapshot and vdisk you want to clone and attach to this instance. c. Click Clone and Attach to Instance. See the topic, Cloning and attaching a vdisk, for more information about cloning and attaching vdisks. Resizing a vdisk 1. On the page of a selected application, click the Profile tab. 2. In the Storage section, in the list of vdisks, click beside the vdisk you want to resize and click Edit Size. 3. Enter the new size of the disk. Note: The new size must be larger than the current size. Page 75

76 4. Click Save. After resizing a disk for Windows 1. Console into the Windows instance, and Open Disk Management. The disk that was resized will show Unallocated space. 2. To create a new volume in the Unallocated space, do the following a. Right-click the Unallocated space and select New Simple Volume. b. Click Next and enter the volume size c. Click Next and assign a drive letter to the new volume. d. Click Next, select formatting options and assign a volume label. e. Click Next and Finish. 3. To extend the current volume into the Unallocated space: a. Right-click the volume and select Extend Volume. b. Click Next and enter the amount of space to add to the volume. c. Click Next and Finish. After resizing a disk for Linux 1. Console into the Linux instance, and do the following: 2. To create a new ext4 partition /dev/sdb2 on disk /dev/sdb: a. Check how many drives are installed: fdisk -l grep '^Disk' b. Add a partition to the disk: fdisk /dev/sdb c. Type p to use the partitioner, n to create a new partition, and select your preferred settings when prompted. d. Type w to save the partition. e. Format the disk: mkfs.ext4 /dev/sdb2 f. Create a mount point and mount the partition: mkdir /disk2 mount /dev/sdb2 /disk2 g. Edit fstab so the new partition automatically mounts to /disk2 on reboot: vi /etc/fstab h. Enter ESC and type i to edit the file. i. Add a line: /dev/sdb2 /disk2 ext4 defaults 1 2 j. Type :wq to write the changes and save the file. 3. To expand an existing, ext4 partition /dev/sdc1 on disk /dev/sdc that's mounted to /disk3: a. Unmount the partition: umount /disk3 b. Expand the partition by deleting and recreating it without deleting the data: fdisk /dev/sdc Page 76

77 c. Print the current partitions: p d. Delete the partition: d e. If multiple partitions are present, select the desired partition when prompted. f. Recreate the partition: Type p to use the partitioner, n to create a new partition, and select your preferred settings when prompted. g. Type w to save the partition. h. Expand the ext4 filesystem: resize2fs /dev/sdc1 Taking a snapshot See the topic, Managing application instance snapshots. Creating a template See the topic, Managing application templates. Inserting guest tools and switching to compatibility/enhanced mode See the topics, Understanding guest agents and instance modes, and, Inserting guest agent tools. Understanding guest agents and instance modes Two instance modes are available: compatibility and enhanced. An application instance can boot in either of these two modes. Instance modes are based primarily on whether you have guest tools installed or not: Enhanced mode is for instances with guest agent tools installed, and Compatibility mode is for instances without guest agent tools installed. Application instances can run without a Cloudistics guest agent and thus you can deploy a wide range of OS types, including virtual appliances like NetScaler. If you decide to create your instance from an ISO or a VM you imported to template (for example, VMDK) it will not have guest tools installed. If you do not want to install guest tools, there are a few things you must know about Compatibility mode. Storage disks are exported as SATA devices instead of SCSI vnics are exported using intel e1000 drivers Snapshots will be crash-consistent instead of application-consistent Memory cannot be increased while the instance is running The Disk Latency graph will not be available When the Cloudistics guest agent is installed and up to date (that is, Enhanced mode), the following features are available for the instance: Cleanly shut down the instance from the Ignite GUI Page 77

78 Cleanly restart the instance from the Ignite GUI Take a snapshot of the running instance External network connectivity for the instance when using virtual networking Also, in Enhanced Mode: The instance will boot if Virtio drivers are already installed. Storage disks are exported to the VM as virtio iscsi devices. vnics are exported to the VM as virtio vnics. You can run an instance in this mode without installing the Cloudistics guest agent. The following capabilities are available in this mode if the guest agent is installed: Snapshots will be application consistent if the guest OS supports freeze and thaw. (They will be crashconsistent if the guest agent is not installed or if the guest OS does not support freeze and thaw.) CPU and Memory statistics as seen from within the guest will be available Docker Volume Driver API support is available. Inserting guest agent tools Note: Guest agent tools are optional. If guest agent tools are not installed, snapshots will be crashconsistent instead of application-consistent, and latency statistics will not be available. For more information about guest agent tools and application instance modes, see the topic, Understanding guest agents and instance modes. You can insert guest agent tools as follows: Insert guest agent tools in the Ignite management console. Install the guest agent tools in your OS. To install the guest agent tools in Windows see this topic. To install the guest agent tools in Linux, see this topic. Note: Guest agent tools can be inserted once the instance starts in an environment running Cloudistics version 3.3 or later. If the instance has been running since before version 3.3, shut down the instance, and then start it (instances that are already shut down can just be started). You can then insert the guest tools. Installing or updating a guest agent tools - Windows Follow these steps to install the guest agent tools for Windows. 1. Insert guest agent tools in the Ignite management console by clicking the Actions button for the instance and selecting Insert guest tools Page 78

79 Note: If the guest tools were inserted in an earlier version and weren't ejected, eject the guest tools then insert guest tools to insert the new version. 2. Insert the Open the guest tools ISO by double-clicking the CDROM in Windows explorer. Note: For instances that already have an older Cloudistics guest agent installed, including new instances created from Cloudistics Marketplace templates, you can skip the Install the drivers section below, and proceed directly to the Install the guest agent section. Installing the drivers 1. Open Device Manager. 2. Expand Other Devices. You will see "unknown" devices marked with a yellow icon: these are the devices that need to be installed. SCSI Controller will always be present, but other drivers may not show up in this stage 3. Right-click each "unknown" device and click Install or Update driver software. Select the option to browse your computer for the software, click Browse, navigate to the CD-ROM, and select the virtio-win folder. Select the checkbox to include sub-folders, and click OK. 4. In the Ignite management console, switch to Enhanced mode for the application instance. a. Select the instance and shut it down. b. On the instance's Profile tab, click Switch to enhanced mode in the Instance Mode section, or click Actions > Switch to enhanced mode. c. Start the instance. 5. Reopen Device Manager and expand Other Devices. There may be new "unknown" devices marked with a yellow icon. Right-click each "unknown" device and click Install or Update driver software. Select the option to browse your computer for the software, click Browse, navigate to the CD ROM and select the virtio-win folder. Select the checkbox to include sub-folders, and click OK. Installing the guest agent tools 1. Navigate to the root of the CDROM drive. 2. Right-click the cloudistics64-{version} Windows Installer Package (cloudistics64-{version}.msi) for 64-bit instances and cloudistics32-{version} Windows Installer Package (cloudistics32-{version}.msi) for 32-bit instances (where {version} is the guest agent version number), and click Install. 3. If any COM+ errors appear during the install, click Ignore. 4. If your instance is still in Compatibility mode, in the Ignite management console: a. Shut down the instance. b. On the instance's Profile tab, click Switch to enhanced mode in the Instance Mode section, or click Actions > Switch to enhanced mode. Page 79

80 c. Start the instance. Installing or updating a guest agent Linux 1. Insert guest agent tools in the Ignite management console by clicking the Actions button for the instance and selecting Insert guest tools Note: If the guest tools were inserted on an earlier version and weren't ejected, eject the guest tools then insert guest tools to insert the new version. 2. Log into the instance. 3. Mount the guest agent tools: mkdir /mnt/cdrom mount /dev/sr0 /mnt/cdrom 4. Install the guest agent tools for CentOS or RHEL: yum localinstall /mnt/cdrom/cloudistics-guest-agent-{version}-1.x86_64.rpm -y replacing {version} with the guest agent version number 5. Install the guest agent tools for Ubuntu: dpkg -i /mnt/cdrom/cloudistics-guest-agent_{version}.deb replacing {version} with the guest agent version number Creating an application instance from ISO Follow these steps to create an application instance from ISO. 1. On the Application Instances page, click New Instance and then select Install from installer. 2. Enter a name for your instance, and configure the required settings. Note that most users should select Compatibility mode. You should only select Enhanced mode if the ISO was created with VirtIO drivers installed. 3. Next to Installer click Select Installer. 4. Enter the information for your NFS or CIFS share. a. Set up an NFS or CIFS share and copy the ISO to it. The share must be accessible by the compute nodes. Note that shares with special characters in the name are not currently supported. Page 80

81 b. On the Nodes page, the compute nodes' IPs will be listed. Since the compute nodes' IPs are assigned via DHCP, not statically, they are subject to change. If only those IPs are given access to the share, or if they change, additional access will have to be granted. Because of this, Cloudistics recommends granting the whole subnet access to the share. c. On the Hardware page, click Show Network Settings; the IP range under Router Physical Network is the range of IPs that could be assigned to the compute nodes. Granting access to the share from this range will allow consistent access to the share from Ignite. 5. Click Create Instance. 6. After the instance is created, open a console into it and complete the installation wizard. 7. Force-shutdown the instance or shut it down from within the instance's OS. This will remove the installation ISO from the instance on the next boot. 8. Start the instance. 9. To install guest tools and switch to Enhanced mode, see these topics for Windows and Linux. Note: For Linux, you might need to install the ACPI daemon to perform soft shutdown/restart. Cloning and attaching a vdisk Why cloning and attaching a vdisk is useful In IT, there are countless situations in which it is useful to create copies of your production data: to perform tasks such as malware analysis, application development and testing (dev/test), user acceptance testing (UAT), analytics, data protection, archiving and provisioning of new instances. It is not uncommon to see 7-10 copies of production data in enterprises today. Unfortunately, traditional copies do not come for free. They consume significant storage, server and network resources, and valuable administrator time and energy. As a result, many IT shops make do with fewer, and less up-to-date copies than they really need. The problem is so significant that entire companies have been created to attack this problem of copy data management. Cloudistics' Clone and Attach vdisk feature allows IT administrators to support many data protection and availability use cases directly within the Cloudistics platform, all from only one golden copy of your production data. The application-consistent clones can be created in seconds, are space efficient and available for quick access while also being deduplicated and compressed for backup or long-term retention. Clones allow you to access a single, point-in-time copy of your primary data. The basic operation is Clone and Attach which allows a consistent copy to be made and immediately attached to a new application instance that needs that copy. Clones may also be deleted once their need is over. Page 81

82 Understanding the clone operation benefits Runs Fast - Unlike today's large, multi-tb drive copies which can take many minutes or hours to complete, Cloudistics' Clone and Attach technology allows even the largest disks to be cloned in a matter of seconds. Reduces Costs - A clone uses a small amount of space for metadata, and then only consumes additional space as data is changed or added. As a result, Cloudistics' Clone and Attach feature can cut the storage you need for dev/test or virtual environments by 50% or more. Supports Business Continuity - Cloudistics' Clone and Attach makes it possible to clone and fully test your DR processes, or use your DR environment for dev/test without interfering with ongoing replication. You simply clone your DR copies and do dev/test on the clones. Improves quality of dev/test - Make as many copies of your full production data as you need. If there are 20 users, give each user their own clone if you wish. If a test corrupts the data, start again in seconds. Developers and test engineers spend less time waiting for access to datasets and more time doing productive work. Accelerates server and virtual desktop provisioning - Deploy tens or hundreds of new instances in minutes with only a small incremental increase in storage directly from our application templates. An important use case for clones is malware analysis in a VDI environment. By using the Cloudistics API, active instances' disks can be cloned and attached to an instance with malware software installed that does scanning for vulnerabilities. The clone can then be deleted. This is accomplished during normal business operations and without disrupting the production instance. How cloning and attaching a vdisk works in Cloudistics Clone disks have all of the capabilities of any other disk, including the ability to be resized, be deduplicated and compressed, be the source of a snapshot, or be cloned. The technology that makes this all possible is integral to how Cloudistics manages storage. Snapshots simply make a copy of the metadata associated with a disk. As data is changed in the base disk, the original blocks stay associated with the snapshot rather than getting marked for reuse. All metadata updates that occur are just pointer changes, so they are very fast. The snapshots are immutable; a clone is simply a snapshot that is writable. You can think of a clone disk as a transparent, writable layer in front of a snapshot. Because a clone disk is writable, it needs disk space to store data that is written to the clone. A snapshot simply links to existing data that was overwritten in the base disk; a clone stores the data written to it on disk and then links to the new data as well. The disk space associated with the snapshot and the clone is accounted for separately from the data in the base disk. Page 82

83 Figure 26 How cloning and attaching a vdisk works in Cloudistics When a clone disk is created, it requires an application-consistent snapshot of the disk to use as a base. The clone disk gets a copy of the snapshot metadata and then updates its metadata as the cloned disk is created. Creating the clone takes just a few seconds because the copied metadata is very small compared to the actual data. All changes to the cloned disk are copy-on-write operations and contain only the changed blocks. The original active disk can change independently of the clone disk because the clone is based on the contents of its base snapshot. Since snapshots are read-only, they can be reused as the base for multiple clone disks. Space is used very efficiently because the only new disk space used is associated either with the small amounts of metadata and/or unique writes to the clone disk. Cloning and attaching a vdisk A vdisk can be cloned from local and disaster recovery (DR) snapshots and attached to an application instance in the same virtual datacenter and storage pool as the snapshot. A vdisk can be cloned and attached as many times as needed. Each clone disk is then independent of other clone disks and of the original snapshot; that is, when you make a change to a clone disk, other clone disks and the original snapshot remain unchanged. Attaching a vdisk to an instance from its own snapshot, Method 1 1. Select Add vdisk from the Actions drop-down menu or click Add vdisk in the Storage section of the instance's Profile tab. 2. Click Add vdisk clone. 3. Select the snapshot and vdisk you wish to clone and attach to this instance. 4. Click Clone and Attach to Instance. Attaching a vdisk to an instance from its own snapshot, Method 2 1. On the instance's Local Snapshots page, click... beside the snapshot and click View vdisks. 2. Click... beside the vdisk you want to clone and select Clone and attach to this instance. 3. Click Clone and Attach to this Instance. Page 83

84 Attaching a vdisk to an instance from another instance's snapshot 1. On the instance's Local Snapshots or Disaster Recovery page, click... beside the snapshot and click View vdisks. 2. Click... beside the vdisk you wish to clone and select Clone and attach to. 3. Select the instance you want to attach the vdisk to, and click Clone and Attach to Instance. 4. After adding the disk in Windows, in the instance's console, open Disk Management, right-click the newly added disk and select Online. 5. After adding the disk in Linux, in the instance's console, do the following: a. Run lsblk to verify that the new, clone disk shows up. (In this example, the new, clone disk is /dev/sdb with partition /dev/sdb1, which has an ext4 filesystem.) b. Create a mount point: mkdir /disk1 c. Mount the clone disk to the mount point: mount /dev/sdb1 /disk1 d. Edit fstab so that the new drive automatically mounts to /disk1 on reboot: i. Type vi /etc/fstab ii. Press ESC and type i to edit the file. iii. Add a line: /dev/sdb1 /disk1 ext4 defaults 1 2 iv. Type :wq to write the changes and save the file. Deleting a clone disk Note: This action deletes the clone vdisk and leaves the original snapshot intact. If changes were made on the clone disk, they will be lost when this disk is deleted. 1. On the Profile tab of an instance, click... beside the clone vdisk and select Delete disk. 2. Click Delete vdisk. Setting a time zone for application instances You can set the time zone in a couple of ways: in Ignite or within the application instance s operating system. Note: Linux instances should use Default or the UTC timezone, rather than a custom local timezone. Follow these steps to set a time zone for an application instance in Ignite. 1. Shut down the application instance for which you want to set a time zone. 2. On the instance's Profile tab, click Actions > Edit timezone configuration. Page 84

85 Note: Default sets the time zone to match the time zone of the compute node. See this topic for information about customizing the time zone for a compute node. 3. Select Custom, the Time zone Area, and the Time zone Location. 4. Click Save, and then start the instance. 5. Set the time zone within the instance s operating system. Configuring NetScaler If you will be using a NetScaler instance as an NFV, you need to configure NetScaler. You need to configure networking for vnic0, which will include giving it an IP in the network and configuring other settings accordingly. Prerequisites for configuring NetScaler for the first time Make sure you obtain the following information before set up: Proper gateway (if it is on a bridged network, it will need to be the gateway IP on that VLAN) Usable IP within your network range Netmask Subnetmask This can done in the command line interface (CLI). set ns config -ipaddress<ipaddress> -netmask<subnetmask> add ns ip<ipaddress> <subnetmask> -type<type> add route<network> <netmask> <gateway> set system user <username> -password save ns config reboot Note: You can ping from "shell" to test this configuration. Example: set ns config -ipaddress netmask add ns ip type snip add route set system user nsroot -password Enter password: ***** Confirm password: ***** save ns config reboot Page 85

86 Working with application templates Refer to the following topics: What is an application template? Viewing application templates Managing application templates Allowing access to a share for importing a VM What is an application template? An application template is a gold master sys-prepped image of an application instance. Templates facilitate the easy and instantaneous creation of application instances; and, a single template can be used to create any number of instances. In Ignite, you can create application templates in several ways: Create a new template from an existing application instance. Import a VM from a.vmdk,.vhd,.vhdx,.img,.ova,.qcow2,.qed,.raw, or.vdi, and save it as a template. Note: Cleanly shut down the source VM before importing the disk file(s). Download a template from the Cloudistics Application Marketplace. What is the Cloudistics application marketplace? The Cloudistics Application Marketplace contains an ever-growing collection of application templates, all of which are free to download. You can access the marketplace by clicking Applications > Marketplace in the side menu in Ignite. Page 86

87 Note: To use the marketplace, the Router Physical Network range must have access to s3.amazonaws.com TCP ports 80 and 443. What is the template store? Templates, whether created internally or downloaded from the marketplace, are stored in the template store. The template store is a portion of the storage pool; as such, it is a local, private storage space that can only be accessed by users within your organization. Any virtual datacenter in your organization with access to a storage pool will have access to all templates held in that storage pool's template store. Viewing application templates To view a list of your application templates, click Applications > Templates in the side menu in Ignite. The Application Templates overview page contains a list of all templates saved in the template store. Each template is listed by name with a description (if any), size, cost, creator, and the date on which the template was last updated. Managing application templates To perform an action on a specific application template, in the Ignite management console, open the Application Templates overview page. Action buttons are located to the right of each template in the list. Creating a new instance 1. Click New Instance to the right of the template. 2. Enter a name (required) and a description (optional) for the new application instance. 3. If you want to create more than one instance, enter the number of instances in the box provided. 4. Select the virtual datacenter where the instance should reside. 5. Select the migration zone where the new instance should reside. 6. If the instance requires a specific compute category or tag, expand Compute Constraints and add the category and tags. 7. Select a storage pool for the application disk(s). 8. Select a networking mode for the instance. If you choose virtual networking, select a virtual network from the drop-down menu. 9. If you do not want the instance to start upon creation, clear the checkbox; start upon creation is enabled by default. 10. Click Create Instance. Page 87

88 Editing a template 1. Click the ellipsis button (...) to the right of the template and select Edit from the drop-down menu. 2. Adjust the template properties as needed. 3. Click Save Changes. Duplicating a template Click the ellipsis button (...) to the right of the template you want to duplicate, and select Duplicate from the drop-down menu. Ignite immediately creates a duplicate copy of the template and adds it to the template store. Deleting a template CAUTION: This action cannot be undone. 1. Click the ellipsis button (...) to the right of the template you want to delete, and select Delete from the dropdown menu. 2. In the confirmation dialog box, click Delete. Allowing access to a share for importing a VM Follow these steps to allow access to a share for importing a VM. 1. To import a.vmdk/.vhd/.vhdx/.img/.ova/.qcow2/.qed/.raw/.vdi via an NFS or CIFS share, on the Templates page, click Import VM. Note: Shares with special characters in the name are not currently supported. 2. On the Hardware page, under the storage block, click Show Details to view the storage block's Virtual IP. On the Hardware page, click each storage node in turn to see its IP address. Storage nodes must be given access to the share for Ignite to view and use the files on it. That is, the storage node's regular IPs and the storage block's virtual IP need access. Since the storage node IPs and the storage block virtual IP are assigned via DHCP, not statically, they are subject to change. If only those IPs are given access to the share, if they change, additional access will have to be granted. Because of this, Cloudistics recommends granting the whole subnet access to the share. On the Hardware page, click Show Network Settings. The IP range under Router Physical Network is the range of IPs that could be assigned to the storage nodes and virtual IP. Granting access to the share from this range will allow consistent access to the share from Ignite. Page 88

89 Page 89

90 Working with application snapshots and disaster recovery Refer to the following topics: What is an application snapshot? Managing application snapshots Managing disaster recovery Using Rapid Recovery What is an application snapshot? An application snapshot is an application-consistent copy of an application instance taken at a specific point in time. Snapshots do not count towards an instance's provisioned storage. You can revert the current instance of an application to an older snapshot at any time without losing subsequent snapshots. You can then make the reversion permanent, which will remove subsequent local snapshots, but the DR snapshots will be retained. In the Ignite management console, a snapshot tree is maintained. The snapshot tree displays the creation date of each snapshot and branch, as well as the size of each snapshot. You can view this tree in the Application Instance: Local Snapshots tab. Managing application snapshots In the Ignite management console, you can take a snapshot of the current instance, revert to a snapshot, or perform maintenance on existing snapshots within your snapshot tree. Page 90

91 To perform an action on a specific application snapshot, select the instance to view its details, and, on the Local Snapshot tab, click the button to the right of the snapshot you want to manage. Taking snapshots 1. At the top of the Local Snapshots tab, click Take snapshot. 2. Enter a name for the snapshot. This name must be unique across all snapshots for this instance. 3. Click Take Snapshot. Reverting a snapshot to an earlier version Note: The current state of the application instance will be lost when you revert to a snapshot. All changes in the instance since the last snapshot will be lost. To save the current state of the instance, take a snapshot before reverting. 1. Click Revert next to the snapshot you want to revert. 2. Click Revert. This will put it in a temporary revert state and no snapshots will be deleted. 3. To return to the latest snapshot, click Return to Latest. No snapshots will be deleted, and the instance will return to the state it was in when the snapshot was taken. 4. To make the reversion permanent, click Make this revert permanent at the top of the page, or Make Permanent beside the snapshot it was reverted to. This will delete the snapshots after this point on the local side, but retain the snapshots on the DR side. Start a snapshot as a new instance 1. Click the ellipsis button ("...") next to the snapshot you want to start as a new instance, and then select New instance from the drop-down menu. 2. Configure settings for your new instance and click Create instance. This creates a new instance from the state the original instance was in when the snapshot was taken. Renaming a snapshot 1. Click the ellipsis button (...) next to the snapshot you want to rename, and then select Rename from the drop-down menu. 2. Enter a new name for the snapshot. This name must be unique across all snapshots for this instance. 3. Click Update. If disaster recovery (DR) is enabled, the snapshot name will change in both the local and DR locations. Viewing vdisks This is helpful if you want to clone a vdisk from a DR snapshot. Page 91

92 1. Click the ellipsis button ("...") next to the snapshot for which you want to view vdisks, and then select View vdisks from the drop-down menu. 2. From here you can clone and attach vdisks by following these steps in this topic, Cloning and attaching a vdisk. Deleting a snapshot CAUTION: Deleting a snapshot cannot be undone. 1. Click the ellipsis button (...) next to the snapshot you want to delete, and then select Delete from the dropdown menu. 2. For manual snapshots, you can select Delete all DR copies of this snapshot to remove the snapshot from the DR side as well as the local side. 3. When you are sure you want to delete it, click Delete Snapshot. Configuring automatic snapshot settings 1. Click the ellipsis button ("...") next to Take Snapshot and select Automatic snapshot settings. 2. Select how often you want automatic snapshots to be taken. 3. Enter the number of local automatic snapshots to retain. Once this number has been reached, the subsequent automatic snapshot will cause the oldest automatic snapshot to be removed. 4. Click Update. Deleting all local snapshots 1. Click the ellipsis button ("...") next to Take Snapshot and select Delete all local snapshots. 2. Click Delete All Snapshots. This step deletes all local snapshots but does not affect DR snapshots. Turning off automatic snapshots 1. Click the ellipsis button ("...") next to Take Snapshot and select Turn off automatic snapshots. Page 92

93 2. Select whether you want to keep DR snapshots and maintain the retention policy, or delete all DR snapshots. 3. Click Turn Off Automatic Snapshots. This step also turns off DR. Invoking Pre- and Post-snapshot scripts You can specify pre- and post- snapshot scripts that can be invoked in an application instance whenever a snapshot is taken. The scripts must be created with the following names and locations: Windows pre-snapshot script: C:\Program Files\Cloudistics\Hooks\PreSnapshot.bat Windows post-snapshot script: C:\Program Files\Cloudistics\Hooks\PostSnapshot.bat Linux pre-snapshot script: /etc/cloudistics/hooks/pre-snapshot Linux post-snapshot script: /etc/cloudistics/hooks/post-snaphot Managing Disaster Recovery Flex DR is a setup and ready-to-use disaster recovery that is the industry s simplest within three clicks. Flex DR includes manual and automatic snapshots with marketplace integration, stateless revert, site-to-site replication and tiered snapshot retention. Requirements Two storage pools, both connected to the same virtual datacenter A network route between storage blocks on each site The network route must be set up so that the storage node IPs and the storage block virtual IPs can communicate. To view the storage block's Virtual IP, on the Hardware page under the storage block, click Show Details. On the Hardware page, click each storage node in turn to see the IP address. Since the storage node IPs and the storage block virtual IP are assigned via DHCP, not statically, they are subject to change. If the network route is only set up between the current IPs and those IPs change, DR will fail. Because of this, Cloudistics recommends setting up the route between the Ignite physical subnets. On the Hardware page under each router, click Show Network Settings; the IP Range under Router Physical Network is the range of IPs that should be used when creating network routes. That is, the Router Physical Network IP Range on one site should have a network route to the Router Physical Network IP Range on the other site. Setting up disaster recovery 1. In ignite, navigate to the Application Instances page, and click the application instance for which DR is being setup. Page 93

94 2. On the Disaster Recovery tab, click Turn on Disaster Recovery, or click the ellipsis button ("...") next to the 3. desired snapshot, and select Turn on Disaster Recovery from the drop-down menu. Figure 27 Disaster recovery settings Understanding disaster recovery settings The first field sets the interval at which snapshots of the instance are automatically taken. Once the number of local automatic snapshots to retain is reached, the oldest automatic snapshot will be removed from local storage, but not disaster recovery. Manual snapshots are not affected by retention policies. The Disaster Recovery Settings set the retention policy for automatic snapshots on the DR storage. Page 94

95 The retention policy fields are not additive. For example, let's use the values from the example settings screen above and assume the first snapshot is taken on 1/1/2017. Each week is assumed to be 7 days, each month is assumed to be 30 days and each year is assumed to be the calendar year, so this example retention policy is set to keep weekly snapshots for 182 days (26 weeks) and monthly snapshots for 180 days (6 months). Since the weekly snapshots are retained longer than the monthly snapshots, weekly snapshots are never rolled into monthly snapshots, so essentially the monthly snapshots interval is ignored. For the first two weeks (1/1/17-1/15/17) all the hourly snapshots are retained on the DR storage pool. After two weeks have passed, the first day's automatic snapshots will be combined into one daily snapshot which will have the data from the newest snapshot of that day. Afer 30 days or more have passed, the oldest daily snapshots will be rolled into weekly snapshots. On 12/31/17 there will be one yearly snapshot 1/1/17, weekly snapshots 7/2/17-11/30/17, daily snapshots 12/1/17-12/16/17, and hourly snapshots 12/17/17-12/30/17. The oldest monthly snapshot would have been 7/4/17 if fewer weekly snapshots were retained. On 1/1/18 the weekly snapshot from 7/2/17 will be older than 26 weeks, so it will be rolled into the next oldest weekly snapshot which was on 7/9/17. There will be one yearly snapshot 1/1/17, weekly snapshots 7/9/17-12/1/17, daily snapshots 12/2/17-12/17/17, and hourly snapshots 12/18/17-12/31/17. On 1/2/18 the snapshot from 1/1/17 will have exceeded the yearly snapshot interval, so it will be rolled into the next oldest snapshot (7/9/17) which will be the new yearly snapshot. There will be one yearly snapshot 7/9/17, weekly snapshots 7/16/17-12/2/17, daily snapshots 12/3/17-12/18/17, and hourly snapshots 12/19/17-1/1/18. Working with DR snapshots actions Starting a snapshot as a new instance 1. Click New Instance next to the DR snapshot from which you want to create a new instance. 2. Configure settings for your new instance and click Create instance. This step creates a new instance from the state the original instance was in when the snapshot was taken. Page 95

96 Viewing vdisks This is helpful if you want to clone a vdisk from a DR snapshot. 3. Click the ellipsis button ("...") next to the snapshot for which you want to view vdisks, and then select View vdisks from the drop-down menu. Page 96

97 4. From here you can clone and attach vdisks by following these steps in this topic, Cloning and attaching a vdisk. Deleting a snapshot Caution: Deleting a snapshot cannot be undone. Automatic snapshots can only be deleted if disaster recovery is turned off; manual snapshots can only be deleted if disaster recovery is turned off and/or the snapshot no longer exists in the source storage pool. 1. Click the ellipsis button ("...") next to the snapshot you want to delete, and then select Delete from the drop-down menu. 2. In the dialog box, click Delete Snapshot. Managing DR snapshot settings Editing the DR retention policy 1. Click the ellipsis button ("...") at the top of the Disaster Recovery tab and select DR retention policy. 2. Modify the settings for the retention policy as needed, and click Update. Page 97

98 Changing the DR location 1. Click the ellipsis button ("...") at the top of the Disaster Recovery tab and select Change DR location. 2. Select the new location. If you have more than two storage pools, you can select a different storage pool to be the DR location. 3. Click Update. Deleting all DR snapshots Caution: Deleting a snapshot cannot be undone. You can only delete all DR snapshots if DR is turned off. 1. Click the ellipsis button ("...") at the top of the Disaster Recovery tab and select Delete all snapshots. 2. In the dialog box, click Delete All DR Snapshots. Note that this does not affect local snapshots. Page 98

99 Turning off DR 1. Click the ellipsis button ("...") at the top of the Disaster Recovery tab and select Turn off disaster recovery. 2. Select whether you want to keep DR snapshots until they are automatically deleted by your configured DR retention policy, or delete all of the DR snapshots now. 3. Click Turn Off Disaster Recovery. Beside each DR snapshot, you can select "New Instance" to create a new instance from the snapshot. Configure settings for your new instance and click "Create instance". This will create a new instance from the state the original instance was in when the snapshot was taken. Manual DR snapshots can be deleted by clicking the ellipsis button ("...") next to the desired snapshot, then selecting "Delete". Click the ellipsis button ("...") at the top of the Disaster Recovery tab and select "DR retention policy" to change the retention policy settings. If you have more than two storage pools, you can click the ellipsis button ("...") at the top of the Disaster Recovery tab, select "Change DR location", and select a different storage pool to be the DR location. If DR is turned off, all DR snapshots can be deleted by clicking the ellipsis button ("...") at the top of the Disaster Recovery tab and selecting "Delete all snapshots". Page 99

100 To turn off DR, click the ellipsis button ("...") at the top of the Disaster Recovery tab and select "Turn off disaster recovery". Select whether you want to keep DR snapshots and maintain the retention policy, or delete all DR snapshots; then click "Turn Off Disaster Recovery". Restoring a file from a snapshot A file can be recovered from a snapshot in one of three ways as described below. Method 1 Follow these steps to clone and attach a vdisk from a snapshot that contains the file to the destination instance. 1. In the instance's console, copy the file from the vdisk clone to the desired location. 2. On the instance's Profile page, click... beside the clone vdisk and select Delete disk. 3. In the warning dialog box, click Delete vdisk. This action deletes the clone vdisk and leaves the original snapshot intact. If changes were made on the clone disk, they will be lost when this disk is deleted. Method 2 Create a new instance from the snapshot containing the file, and copy the file from there to the desired location by following these steps. 1. Click... beside the snapshot, and then select New instance from the drop-down menu. 2. Configure the settings for your new instance, and click Create instance. This step creates a new instance from the state that the original instance was in when the snapshot was taken. Page 100

101 3. Copy the file from the new instance to the new location of the file. 4. For the new instance created from the snapshot, select Delete on the Actions drop-down menu, and then click Delete Instance. Method 3 Follow these steps to revert the instance to the snapshot containing the file, copy the file to another location, return the instance to its latest snapshot, and then copy the file from the other location to the instance. 1. On the instance's Local Snapshots tab, click Revert next to the snapshot that contains the file. Page 101

102 2. Select the checkbox Take a snapshot of the current instance before reverting, and then click Revert. This step puts the instance in a temporary revert state, and no snapshots will be deleted. 3. In the instance's console, copy the file to a location external to the instance. 4. Click Return to Latest. No snapshots are deleted, and the instance returns to the state it was in when the last snapshot was taken. 5. Copy the file from the external location onto the target instance. Page 102

103 Using Rapid Recovery Backing up an Ignite Server 2012 R2 with Rapid Recovery When backing up an Ignite Server 2012 R2 instance with Rapid Recovery, the Cloudistics service should be stopped before taking a backup and started again afterwards. To automate this, use Rapid Recovery pre- and post-transfer scripts ( On the 2012 R2 agent, open notepad and enter Stop-Service cloudistics; save the file as C:\Users\All Users\AppRecovery\Scripts\PreTransferScript.ps1 On the 2012 R2 agent, open notepad and enter Start-Service cloudistics; save the file as C:\Users\All Users\AppRecovery\Scripts\PostTransferScript.ps1 Note: Only one backup solution should be used for an instance. If an instance is being backed up by Rapid Recovery, Ignite automatic snapshots should be turned off. Performing a Rapid Recovery restore Follow these steps to inject drivers into the Universal Recovery Console to perform a Rapid Recovery bare metal restore. 1. Create a new instance from installer, using the Rapid Recovery Boot CD. Make sure to allocate the same amount of compute and storage resources as the original instance. 2. Open the console to the new instance and launch the Universal Recovery Console. 3. Load the network driver. 4. Click Force Load. In the popup window, expand the volume labeled virtio-win Open the NetKVM folder, then open the folder whose name matches the OS that's being restored, and then open the folder that matches the architecture that's being restored (amd64 for 64-bit, x86 for 32-bit). Select netkvm.inf and click OK. 6. Load the storage driver. 7. Click Force Load. In the popup window, expand the volume labeled virtio-win Open the vioscsi folder, then open the folder whose name matches the OS that's being restored, and then open the folder that matches the architecture that's being restored (amd64 for 64-bit, x86 for 32-bit). Select vioscsi.inf and click OK. Perform the remaining steps outlined here: Page 103

104 Monitoring and maintaining your environment Refer to the following topics: Monitoring resources and statistics in Ignite Understanding storage activity information Managing users Configuring notification settings Using Docker in Ignite Editing SDN router network settings Setting up PXE Boot in Ignite Configuring Citrix PVS in Ignite Forwarding hardware logs to a remote server Understanding CPU load vs CPU in-use metrics Powering down hardware in Ignite Monitoring resources and statistics in Ignite Ignite provides a convenient dashboard for viewing all of our resource activity. It provides a single-view overview of your infrastructure with a Web representation of your hardware stack as well as Disaster Recovery and virtual data center information. Ignite also provides insight in how your compute and storage resources are allocated. With performance monitoring, you always know what applications are using resources as well as peak times. Ignite gives you full visibility into the applications in your cloud, including OS, network bandwidth, storage performance, latency, and more. Click the Dashboard link in the side menu to open your dashboard to view live statistics and monitor your resources. Page 104

105 Figure 28 Example Ignite dashboard Understanding storage activity information (queue depth, IOPS, and latency) In Ignite, in the storage section of your instances activity page you will see such information as IOPs, and Queue length. This topic describes in more detail the information that is displayed and how you can determine your IO rate and latency. Input/output operations per second (IOPS, pronounced eye-ops) is an input/output performance measurement used to characterize computer storage devices like hard disk drives (HDD), solid state drives (SSD), and storage area networks (SAN). Queue Depth (queue length) is the number of I/O requests (SCSI commands) that can be queued at one time on a storage controller. Each I/O request from the host's initiator HBA to the storage controller's target adapter consumes a queue entry. (Note: this number will vary as it shows how much your system is maxing at that moment) Calculating your IO rate Latency is a measure of how long it takes for a single I/O request to happen from the application s viewpoint. Page 105

106 For example, if your charts show that IOPS are at 160,000 then that's 160k IO occurring in 1 second which makes 1 IO of a second. Using Centos with FIO Marketplace template to determine latency Note: The recommended test to run is a Direct I/O, which means an IO access where you bypass the cache (as buffered can be misleading) You need to stop the automatic FIO that is running and manually run it. 1. Download the Centos with FIO template from the marketplace and open it. 2. Run the command: systemctl stop runfio 3. To check the parameter settings, run the command: Cat rw_test.fio 4. Run the command: /usr/bin/fio /root/rw_test.fio To stop the test at any time you, press the keys CTRL-C. The test will look similar to the following example. In this example, you can see the same information provided on the activity page, including in depth details of your environment. Figure 29 Example of running rw_test.fio Page 106