Dell EMC Data Protection Search

Transcription

1 Dell EMC Data Protection Search Version 1.1.x Installation and Administration Guide REV 05

2 Copyright Dell Inc. or its subsidiaries. All rights reserved. Published May 2017 Dell believes the information in this publication is accurate as of its publication date. The information is subject to change without notice. THE INFORMATION IN THIS PUBLICATION IS PROVIDED AS-IS. DELL MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. USE, COPYING, AND DISTRIBUTION OF ANY DELL SOFTWARE DESCRIBED IN THIS PUBLICATION REQUIRES AN APPLICABLE SOFTWARE LICENSE. Dell, EMC, and other trademarks are trademarks of Dell Inc. or its subsidiaries. Other trademarks may be the property of their respective owners. Published in the USA. Dell EMC Hopkinton, Massachusetts In North America Data Protection Search 1.1.x Installation and Administration Guide

3 CONTENTS Preface 7 Chapter 1 About Data Protection Search 11 Data Protection Search Introduction...12 Accessing only the Search window Accessing the Administration window Product architecture...14 Data Protection Search scalability Elasticsearch...16 Replication Elasticsearch analyzers and tokenization Data Protection Search cluster components...19 Data Protection Search licensing...20 NDMP support in Data Protection Search Chapter 2 Installation 21 Data Protection Search installation planning and considerations Virtual Appliance deployment preparation Data Protection Search node component requirements Deploying the Virtual Appliance on the vcenter server Deploying the vapp from a template...28 Deploying the vapp from VMware vcloud Director...30 Initializing the Data Protection Search environment...31 Basic configuration Advanced configuration...36 Installing Index Data components Installing Worker components Installing additional Worker nodes Installing a self-signed or trusted certificate...42 Update LDAP configuration Updating LDAP configuration in the Data Protection Search Admin UI...44 Updating LDAP configuration in the Data Protection Search Admin installation script Updating the Data Protection Search Admin Group...46 Upgrading to the current release of Data Protection Search...46 Chapter 3 Administration 49 Data Protection Search Administration Web Application...50 Logging in to the Data Protection Search Admin user interface Data Protection Search Admin UI home...51 Data Protection Search dashboard...52 Chapter 4 Sources 55 Sources Adding an Avamar server to Data Protection Search...56 Data Protection Search 1.1.x Installation and Administration Guide 3

4 CONTENTS Default Avamar server limit Adding a NetWorker server to Data Protection Search Connection Limitations considerations Updating an Avamar or NetWorker server...60 Removing a server from Data Protection Search Registering agents manually...61 Chapter 5 Roles 63 Roles UI...64 Adding Index Admins Removing an Index Admin...66 Data Protection Search Admin role Chapter 6 Indexes 67 Indexes view...68 Adding an index Editing an index Removing an index Chapter 7 Collections 71 Collections activities...72 Adding a collection activity Collection activity fields...74 Editing a collection activity...75 Chapter 8 Jobs 77 Jobs overview...78 Jobs views Data Protection Search job types...80 Chapter 9 System 83 Data Protection Search System Worker Nodes...84 CIS Nodes Chapter 10 Options 89 Data Protection Search Options...90 Configuring notifications...91 Chapter 11 Performing Searches 93 Search overview...94 Search performance factors...94 Search by using keywords Search criteria management...97 Using Lucene to search Restore files in Data Protection Search Search results Search criteria management Search performance factors Data Protection Search 1.1.x Installation and Administration Guide

5 CONTENTS Chapter 12 Troubleshooting 107 The Data Protection Search log files Copying log files by using WinSCP Copying log files by using PuTTy Viewing and filtering log files with Data Protection Search log viewer Troubleshooting the Data Protection Search web server Data Protection Search configuration files Troubleshooting web services for collector issues Increasing the maximum memory for the dpworker service Elasticsearch troubleshooting Controlling the Elasticsearch service Viewing Elasticsearch logs Viewing or changing Elasticsearch configuration Monitoring the health of the Elasticsearch cluster Glossary 115 Data Protection Search 1.1.x Installation and Administration Guide 5

6 CONTENTS 6 Data Protection Search 1.1.x Installation and Administration Guide

7 Preface As part of an effort to improve product lines, periodic revisions of software and hardware are released. Therefore, all versions of the software or hardware currently in use might not support some functions that are described in this document. The product release notes provide the most up-to-date information on product features. If a product does not function correctly or does not function as described in this document, contact a technical support professional. This document was accurate at publication time. To ensure that you are using the latest version of this document, go to the Support website at support.emc.com. Revision history This revision history provides a description for each revision of this Installation and Administration guide. Table 1 Data Protection Search Revision History Revision Date Changes 05 May 30, 2017 Updated for Data Protection Search 1.1 SP3, including End User License Agreement (EULA) updates. 04 August 17, 2016 Added Default Avamar server limit section to the Sources chapter. 03 February 17, 2016 Editorial updates. 02 October 23, 2015 Modified the Data Protection Search Upgrade procedure in the Installation section of this guide. 01 September 23, 2015 Initial revision of the Data Protection Search Installation and Administration Guide. Purpose This document describes how to install, configure and use Data Protection Search. Audience This document is intended for the search administrator and index administrator who will be involved in managing Data Protection Search. Related documentation The Data Protection Search documentation set includes the following publications: Data Protection Search Software Compatibility Guide Data Protection Search 1.1.x Installation and Administration Guide 7

8 Preface Data Protection Search Security Configuration Guide Data Protection Search Installation and Administration Guide Data Protection Search Release s Special notice conventions that are used in this document The following conventions are used for special notices: NOTICE Identifies content that warns of potential business or data loss. Contains information that is incidental, but not essential, to the topic. Typographical conventions The following type style conventions are used in this document: Table 2 Style conventions Bold Italic Monospace Used for interface elements that a user specifically selects or clicks, for example, names of buttons, fields, tab names, and menu paths. Also used for the name of a dialog box, page, pane, screen area with title, table label, and window. Used for full titles of publications that are referenced in text. Used for: System code System output, such as an error message or script Pathnames, file names, file name extensions, prompts, and syntax Commands and options Monospace italic Monospace bold Used for variables. Used for user input. [ ] Square brackets enclose optional values. Vertical line indicates alternate selections. The vertical line means or for the alternate selections. { } Braces enclose content that the user must specify, such as x, y, or z.... Ellipses indicate non-essential information that is omitted from the example. You can use the following resources to find more information about this product, obtain support, and provide feedback. Where to find product documentation Where to get support The Support website at provides access to licensing information, product documentation, advisories, and downloads, as well as how-to and 8 Data Protection Search 1.1.x Installation and Administration Guide

9 Preface troubleshooting information. This information may enable you to resolve a product issue before you contact Support. To access a product specific Support page: 1. Go to 2. In the Find a Product by Name box, type a product name, and then select the product from the list that appears. 3. Click the following button: 4. (Optional) To add the product to My Saved Products, in the product specific page, click Add to My Saved Products. Knowledgebase The Knowledgebase contains applicable solutions that you can search for by solution number, for example, , or by keyword. To search the Knowledgebase: 1. Go to 2. Click Advanced Search. The screen refreshes and filter options appear. 3. In the Search Support or Find Service Request by Number box, type a solution number or keywords. 4. (Optional) To limit the search to specific products, type a product name in the Scope by product box, and then select the product from the list that appears. 5. In the Scope by resource list box, select Knowledgebase. The Knowledgebase Advanced Search panel appears. 6. (Optional) Specify other filters or advanced options. 7. Click the following button: Live chat To participate in a live interactive chat with a support agent: 1. Go to 2. Click Chat with a Support Agent. Service requests To obtain in-depth help from Support, submit a service request. To submit a service request: 1. Go to 2. Click Create a Service Request. To create a service request, you must have a valid support agreement. Contact a sales representative for details about obtaining a valid support agreement or with questions about an account. To review an open service request: 1. Go to Data Protection Search 1.1.x Installation and Administration Guide 9

10 Preface 2. Click Manage service requests. Online communities Go to the Community Network at for peer contacts, conversations, and content on product support and solutions. Interactively engage online with customers, partners, and certified professionals for all products. How to provide feedback Feedback helps to improve the accuracy, organization, and overall quality of publications. You can send feedback to 10 Data Protection Search 1.1.x Installation and Administration Guide

11 CHAPTER 1 About Data Protection Search This section includes the following topics: Data Protection Search Introduction Accessing only the Search window Accessing the Administration window...13 Product architecture Elasticsearch Data Protection Search cluster components About Data Protection Search 11

12 About Data Protection Search Data Protection Search Introduction Data Protection Search is a scalable, fault-tolerant, indexing, and search virtual appliance. Data Protection Search provides a powerful way to search backup data across one or more Avamar and NetWorker servers, and then restore or download the search results. Scheduled collection activities are used to gather and index the metadata and/or content of backups, which is then stored within the Data Protection Search node or cluster. Index Data Protection Search indexing provides the following support and capabilities: Processes content from multiple input sources for Avamar and NetWorker servers in Data Protection Suite Supports both metadata only, and full content indexing: Metadata only index is a lightweight search for the name, location, size, and date for files Full content index expands on metadata only index by drilling into the body of files to index the content Leverages Elasticsearch, a scalable, fault-tolerant open source indexing technology Search Data Protection Search provides the following search support and capabilities: A simple, familiar Google like search interface Complex, advanced search queries Cross server and cross platform searches Supports the following actions on search hits: Preview Download Restore to both original and alternate locations Accessing only the Search window The Data Protection Search virtual appliance is accessed as a web-based interface. Depending on your user privileges, you can perform administration or search tasks. To perform searches, you must be assigned one of the following roles: DPSearch Admin Index Admin You must log in to the Data Protection Search UI each time it is opened, or after an inactivity time out of 20 minutes or more. The default inactivity timeout is 20 minutes. To perform searches, complete the following steps. 12 Data Protection Search 1.1.x Installation and Administration Guide

13 About Data Protection Search Procedure 1. In a web browser, type the following: You might be required to acknowledge a browser warning regarding self-signed certificates before continuing. 2. To log in: a. Type the username and password. The default username is Admin. A System Administrator or Application Administrator can assign roles to users. b. Click Sign In. The Search UI opens. Accessing the Administration window The Data Protection Search virtual appliance is accessed as a web-based interface. Depending on your user privileges, you can perform administration or search tasks. To perform administration tasks, you must be assigned the Index Admin role. You must log in to the Data Protection Search UI each time it is opened, or after an inactivity time out of 20 minutes or more. The default inactivity timeout is 20 minutes. To perform administration tasks, complete the following steps. Procedure 1. In a web browser, type the following: For example: You might be required to acknowledge a browser warning regarding self-signed certificates before continuing. 2. To log in: a. Type the username and password. Accessing the Administration window 13

14 About Data Protection Search The default username is Admin. b. Click Sign In. The Administraton UI opens. 3. To open the Administration window, click the following button: The Administration window appears and displays the available menu icons. 4. To open the Search window, click the following button: The Search window appears and displays the search options. Product architecture Data Protection Search is a pre-installed, Linux-based virtual appliance. Additional configuration is required during deployment. Data Protection Search can be deployed as a single node, or multiple nodes can be used to form a fault-tolerant cluster. The following figures illustrate the architectural components of Data Protection Search and the data flow. Figure 1 Data Protection Search architecture 14 Data Protection Search 1.1.x Installation and Administration Guide

15 About Data Protection Search Figure 2 Data Protection Search system The following table lists the components that make up the Data Protection Search architecture. Table 3 Data Protection Search components Component Elasticsearch Common Index Service (CIS) Collector Service Apache Tika NGINX web server DPSearch Admin app DPSearch Search app LDAP Description Highly scalable full-text index and search platform with RESTful web interface, schema-free documents, and built-in replication. An abstraction layer above Elasticsearch that provides the ability for multiple applications to share the same Elasticsearch cluster, enabling cross-platform searches. CIS also provides a security layer in front of Elasticsearch to prevent unauthorized access. Manages scheduled jobs to collect metadata and, optionally, full content of backup files. The collector service leverages connector interfaces to interact with backup platforms. Open source toolkit to extract full-text content and applicationspecific metadata from a wide variety of file types. Open source reverse proxy web server, hosting the web-facing components of Data Protection Search. Web application that serves as the primary interface for Data Protection Search and Index Administrators to manage the configuration of the product. The configuration tasks include managing backup servers, scheduling collection activities, roles, options, and monitoring the system. Web application that serves as the primary interface for Search Admins to search, preview, download, and restore indexed content. Searches may be on basic file or backup metadata, such as keyword, name, type, location, size, and backup server/client, or full-text content of the files. Also supported is visual filtering and analysis, and a powerful text-based query language. All administrators (users and groups) leveraging the Data Protection Search Admin and Search applications are selected from an existing LDAP server, such as Microsoft Active Directory. Product architecture 15

16 About Data Protection Search Table 3 Data Protection Search components (continued) Component Unicorn Description The CIS components of Data Protection Search use Unicorn, an HTTP server for Rack applications. The Unicorn service must be running on the Data Protection Search Index Master node for CIS to work. At the terminal, the status of the service can be checked by using service unicorn status, and if required it can be started with service unicorn start. Data Protection Search scalability Although a single node is supported for the Data Protection Search environment, it is extremely scalable for both Elasticsearch nodes and the Worker nodes. A benefit of a multi-node environment is that you can configure index data, and worker nodes without first knowing the number of each node required. You can add nodes as required. The following image illustrates an example of a multi-node Data Protection Search environment. Figure 3 Multi-node environment Elasticsearch At the core of Data Protection Search is Elasticsearch, an open source, scalable technology that supports indexing, searching, and analytics. Elasticsearch leverages Apache Lucene for its indexes. Elasticsearch does work well on a single node, however to leverage its built in replication functionality, at least two Index Data nodes are required. Additional nodes 16 Data Protection Search 1.1.x Installation and Administration Guide

17 About Data Protection Search improve search performance. The various elements and concepts of an Elasticsearch cluster are described in the following table. Table 4 Elasticsearch cluster concepts and elements Term Cluster Node Description A single node or collection of nodes. An Elasticsearch process running on the JVM. For Data Protection Search, Master node (1 per cluster) and Data nodes are the primary nodes. All indexing a query traffic must pass through the Master node. If this node fails, all activity stops until such time it becomes available. To balance the work load, to avoid failure on the Master node, add extra Data nodes. Shard Primary shard Replica shard A Lucene index. Indexed documents go to the primary shard first. A copy of the primary shard that increases both performance and failover capability. The way Elasticsearch organizes indexed content is similar to that of a relational database, although the terminology differs somewhat. The following table describes the terms used for indexing, and the equivalent relational database terms. Table 5 Elasticsearch index terminology Term Relational database equivalent Description Index Database Similar concept to a relational database with mappings. Physical indexes are implemented as shards. Mapping Schema The schema/definition for the index. Optional, Elasticsearch makes a best guess. Type Table Multiple types can exist within a single index. Document Row A JSON object representing an item composed of key/ value pairs, for example, Name: filename.docx. Field Column For the example, Name: filename.docx, Name is the field. Term Not applicable Term is the value stored in the index. Elasticsearch 17

18 About Data Protection Search Replication Replication in Elasticsearch provides built-in fault-tolerance. In order for replication to work, there must be at least two Indexing nodes deployed. Multiple replicas are supported if there are at least the number of replicas + 1 nodes available. Each index is divided into shards (5 is the default). The shards are distributed over the available Index nodes. For each replica, there is a Replica shard. Replica shards must be stored on a different node from the corresponding Primary shard. For example, an index with five shards, and two replicas has a total of 15 shards. The shards are distributed among the available nodes. If the cluster does not include least three nodes, Data Protection Search marks the cluster and index with a yellow warning state, as it is not possible to create all requested replicas. When indexing occurs, each item is written to at least the primary and replica shards. Therefore, the item must have been written successfully to a majority of the primary and replica shards before returning. In the example above, the item must have been written to the primary shard and one replica before returning. The item is replicated to the final replica later on. When an Elasticsearch node becomes unavailable, replica shards are automatically promoted to replace any unavailable primary shards without required administrator interaction. Elasticsearch analyzers and tokenization When an item is indexed, Elasticsearch converts the various elements into a form that can be stored in the index and used to search later on. The process of breaking up text along word boundaries and removing punctuation is called tokenization. Internally, various filters are used in this process. The groupings of rules that determine which filters and transformations run are called analyzers. If required, Elasticsearch has a number of built in analyzers, and Data Protection Search provides the ability for analyzers to be selected for a particular index Data Protection Search. One example scenario in which an alternate analyzer selection might be appropriate is to apply stemming and/or stop words for full-content indexing. Stemming modifies words to their basic root or stem., and provides the ability for keyword searches to match a broader set of hits. For example, searching for run returns a match if the index includes any of the following: Run Runs Running Runner Stop words prevent unnecessary words from being indexed. Stop words result in smaller indexes, faster queries, and broader matches. For example, indexing "The cat sat on the mat" will only index "cat", "sat", and "mat". Similarly, searching for "The cat sat on the mat" automatically strips out "the" and "on" from the search query. Stemming and stop words are supported for many different languages. Supported Data Protection Search analyzers are listed in the following table. 18 Data Protection Search 1.1.x Installation and Administration Guide

19 About Data Protection Search Table 6 Supported Data Protection Search analyzers Analyzer type Standard (default) Description The standard analyzer: Lower cases Handles text and numeric tokens Tokenizes , and other text strings correctly It is strongly recommended that you use the Standard Data Protection Search Analyzer. Simple The simple analyzer: Lower cases Discards numeric values Splits text into tokens on non-letter characters Whitespace Languages The whitespace analyzer does not support lowercases and splits text into tokens on whitespace characters only. The languages analyzer: Supports stemming as appropriate for the language Removes stop words as appropriate for the language Supports a very large selection of languages Although the default standard analyzers are recommended for the majority of use cases, advanced users can select different analyzers for specific indexes if required. Data Protection Search cluster components Configure a Data Protection Search cluster with multiple nodes or deploy all components on a single node. The three main components of the Data Protection Search cluster environment are: Worker nodes (worker nodes can be restricted to either web server only, or collection service only) Index Master nodes (including CIS) Index Data nodes Extra Worker nodes can be added to improve the speed of collections. Extra Data nodes can be added to: Improve the speed of indexing Improve the speed of queries Store more indexed content Provide replication Data Protection Search cluster components 19

20 About Data Protection Search Data Protection Search licensing Data Protection Search does not require its own license, and is free with a Data Protection Suite license. NDMP support in Data Protection Search NDMP backup indexing for some NAS devices is supported for Avamar and NetWorker in Data Protection Search. The Data Protection Search Compatibility Guide provides details on which NDMP NAS devices are supported for indexing. The following table lists supported NDMP operations in Data Protection Search for Avamar and NetWorker. Table 7 NDMP support Operation Avamar NetWorker Metadata Indexing Yes Yes Full content Indexing Yes No Download Yes No Restore to original client Yes Yes Restore to NDMP client (same type) Restore to NDMP client (different type) Yes No Yes No Restore to file system Yes No 20 Data Protection Search 1.1.x Installation and Administration Guide

21 CHAPTER 2 Installation This section includes the following topics: Data Protection Search installation planning and considerations...22 Virtual Appliance deployment preparation Deploying the Virtual Appliance on the vcenter server...27 Initializing the Data Protection Search environment Installing a self-signed or trusted certificate Update LDAP configuration...43 Upgrading to the current release of Data Protection Search...46 Installation 21

22 Installation Data Protection Search installation planning and considerations Data Protection Search is a virtual appliance that can be deployed to VMware vcenter, or vcloud with one or more VMDKs. How large is the environment When planning the Data Protection Search deployment, there are a number of factors to consider. The number of backup servers in the environment, and more importantly, the total number of clients to be indexed across the servers makes up the environment. The next factor to consider is how many backup records exist on the clients. For example, if a particular client contains 1 million files, and there are daily backups that are retained for 30 days, there may be 30 million backup records for that client. What is being indexed? Data Protection Search only indexes traditional file system backups. The following backup types are ignored: Virtual machines Databases Microsoft modules Snapshots Block based backups Data Protection Search provides flexibility for what is indexed from clients. It can be restricted to certain file types, and files modified within a certain date range. The range of backup dates can be specified, either per client, or for all clients. Processing all historical backups takes longer than only indexing a short range of backups, or day-forward backups. By default, collection activities index backups less than a week old (before the collection) can be modified. To keep the index current, it is recommended that collections are scheduled to run on the same schedule as backups. For example, daily or weekly. Metadata or full content indexing? You can specify the type of indexing for each collection activity. The default indexing method is metadata only, and basic metadata is collected for each file: File name Path Size Date Server operating system Client File location (backup details) Information is available from the Avamar or NetWorker databases, so processing finishes quickly, and backup storage access is not required. The amount of space that is required to store the record in the index is minimal. Full content indexing restores full file data to the Worker node. The data is scanned, and all text strings are extracted and indexed. The text is searchable, and a preview of the file, either text-based or a thumbnail image is stored in the index. Full content 22 Data Protection Search 1.1.x Installation and Administration Guide

23 Installation searches provide an enhanced search experience, at the expense of slower indexing, larger index size, and impact on backup storage performance. It is recommended that metadata only is the primary search method. How many Data Protection Search Appliances are required? It is possible to deploy Data Protection Search as a single, all-in-one node for both indexing, and search. If enough storage is available, an all-in-one node can handle billions of backup records. However, search performance is reduced as the number of records increases. Filtered searches can take a few seconds, or less, but broader searches and visual filters take longer. Adding additional Index Data nodes provides better search performance. The computing, and memory requirement improvement is noticeable for concurrent searches distributed across all nodes. Additional Data Protection Search Worker nodes enable collection activities to process faster, as the work items are distributed across all nodes. Add additional Worker nodes if there is a limited indexing window, and Data Protection Search cannot process all clients in that window. Adding additional Worker nodes to process full content indexing for one Avamar server does not increase performance as the server is the bottleneck, not Data Protection Search. For example, an increase of 50%-70% indexing performance after one extra node, but only 40% for the next node, and decreases with every node added. The initial collection (index) takes longer than subsequent collections. Just as with full and incremental backups, Data Protection Search must initially process all files, and then process new or changed files. Indexing weeks, or months worth of historical backups takes longer than processing current daily backup collections. Are replicas required? Elasticsearch can replicate, and fail over indexes automatically. For one Data Protection Search node without replication, changing the number of replicas to 0 prevents a yellow warning icon from displaying for the Elasticsearch cluster, and indexes on the dashboard. If there are multiple nodes, it is recommended that replica is not set to 0. If there is no replica, and fail over cannot occur. There must be at least replicas plus one Index Master, or Data node in a Data Protection Search cluster. Additional Index Data nodes improve search performance for replication. How much space is required for the indexes? The space that is required for indexes depends on a number of factors: Metadata only or full content indexing For full content indexing, the size and type of files being indexed For metadata indexing, the length of file names and pathnames The amount of duplication (unchanged files appearing in multiple backups) Replication For metadata only, with 80% duplication, each backup record can use approximately 100 bytes to 200 bytes. With zero duplication (all unique files) which is unlikely, the average can be as large as 400 bytes to 600 bytes. Data Protection Search installation planning and considerations 23

24 Installation For full content indexing, the range is wider. Large multimedia files use very little space in the index, while small documents use significantly more. With a typical dataset at 80% duplication, each backup record can use 1 KB to 4 KB. With zero duplication, that average can be as large as 15 KB to 30 KB, or more. Table 8 Disk use based on indexing Metadata only index Backup records 1 billion 5 billion 10 billion 20 billion 100 bytes 93 GB 466 GB 931 GB 1.9 TB 200 bytes 186 GB 931 GB 1.9 TB 3.7 TB Full-content index 1 KB 854 GB 4.8 TB 9.5 TB 19 TB 4 KB 3.8 TB 12 TB 38 TB 76 TB How large will the indexes grow over time? If an activity is setup to recur (daily), Data Protection Search continues to process new backups/save sets as they are created. New or modified items are added to the index, and there are references for items that are unchanged from previous backups. When a backup expires, or is deleted, any items no longer in any backups for that client are removed from the index. Daily garbage collection, and a monthly reconciliation job removes expired, or deleted items. The index size for a particular client initially grows until all current backups (for example, 30 days worth for monthly backup retention) are in the index. Then the index size stabilizes, and growth matches the data growth on that client. Adding more clients increases the size of the index. View the size of the index in its properties. Indexes can be deleted to recover space. How long does a search operation take? The Data Protection Search and Elasticsearch framework is immediate, with searches often taking less than a second. However, as the number of items in an index increases, search speed is often impacted as well, particularly on a single node. Generally, the broader the search is, the greater the impact of a large index is. The broadest possible search is a wildcard (*) search across all indexes. Wildcard searches can complete in a few seconds even with hundreds of millions of indexed items, but as the item count increases to the billions, searches might time out before completing. The recommendation is to use filters to narrow searches as much as possible. A wildcard (*) Search can take 30 s to 60 s in a single node environment with 15 billion backup records (1.5 billion unique files. A search that using type, and client keyword filters can take less than 1 s to complete. Since millions or billions of results are not useful, the use of filters is recommended. If indexing is running during the search, broad searches perform faster with a static index. 24 Data Protection Search 1.1.x Installation and Administration Guide

25 Installation Visual filters require that matching items from the current search are aggregated, and requires large amounts of memory and time. Filters can narrow the results being aggregated to reduce the impact on memory and time. If a visual filter is not completed in a timely manner (10 s), the current results are displayed with a warning message indicating that results are incomplete. Virtual Appliance deployment preparation Before beginning the Data Protection Search Virtual Appliance deployment, create, or collect the necessary groups, users and information. Required information The following information is required for the deployment and configuration of Data Protection Search: Network: Hostname and IP DNS server Gateway Domain name LDAP: Hostname or IP Port Base domain name Username/password User account with the ability to query LDAP Data Protection Search Admin Group The following are the minimum virtual hardware memory requirements for the virtual machine: 32 GB RAM 2 CPUs 500-GB disk A dedicated Worker node, that is not configured for Elasticsearch needs the default 40-GB disk If Elasticsearch is being used (Index Master or Index Data node), it requires an additional disk that varies in size depending on the requirements. LDAP Groups and Users Data Protection Search uses Active Directory for any LDAP services: CIS Admin to administer the Common Indexing Service (CIS), which is the the security and multi-indexing layer that sits on top of Elasticsearch: Used to query LDAP It is helpful to create a user and password for the CIS Admin that does not expire Virtual Appliance deployment preparation 25

26 Installation Create a Data Protection Search Admin Group Create at least one Admin user and add the user to the Data Protection Search Admin Group. A Data Protection Search Admin user can also be an Index Admin and a Search Admin (for evaluation) (Optional) Create the following: Index Admins (users and/or groups) Index Admins manage indexes, and manage and monitor collection activities and jobs. Index Admins receive notifications for those jobs they started. Search Admins (users and/or groups) Each index has a list of Search Admins (users/groups). Only Search Admins belonging to a specified index can search that index. Search Admins are specified Read only, or Full access permissions for a specified index. Members of the Data Protection Search Admin Group are also Index Admins by default. Virtual appliance Data Protection Search is a virtual appliance composed of an OVF and a single VMDK. Convert the deployed virtual machine to a template, and deploy the template as required for Data Protection Search nodes. You can partially configure the template to simplify, and speed up future node deployments: Accept the EULA and then answer the following question: Is this appliance being deployed in China, Taiwan, Hong Kong, or Macau? y(es) or n(o) (Default: no) Setup common networking values, such as domain, DNS, and routing Setup the date/time zone Update/change passwords Data Protection Search node component requirements There are several factors to consider when planning the Data Protection Search cluster. The following table describes the Data Protection Search node requirements. Table 9 Data Protection Search node requirements Node CPU Memory Disk All in one GB+ Disk 1: 40-GB Worker GB 40 GB Index Master GB Disk 1: 40-GB A second disk is required for nodes running as Elasticsearch data nodes, including all-in-one node. Disk 2 is mounted to a directory for Elasticsearch at deployment. 26 Data Protection Search 1.1.x Installation and Administration Guide

27 Installation Table 9 Data Protection Search node requirements (continued) Node CPU Memory Disk Index Data GB Disk 1: 40-GB An Index Master node also serves as an Index Data node. Therefore, a second disk is required. Disk 2 is mounted to a directory for Elasticsearch at deployment. A second disk is required for nodes running as Elasticsearch data nodes. Disk 2 is mounted to a directory for Elasticsearch at deployment. Deploying the Virtual Appliance on the vcenter server The Data Protection Search Virtual Appliance can be deployed to VMware vcenter by following the wizard. Before you begin Review the following sections before deployment: Environment and system requirements Data Protection Search node component requirements Virtual Appliance deployment preparation Procedure 1. From a vsphere client that connects to a VMware vcenter Server with ESX hosts, click File > Deploy OVF Template. A vcenter server is required to deploy the OVF. 2. Browse to the DPSearch.ovf file and click Next. 3. In the Name and Location window, specify a Name and an Inventory Location for the deployed template, and click Next. 4. Select a Host / Cluster on which to run the deployed template and click Next. 5. Select a Resource Pool for which to manage storage, and click Next. 6. In the Storage window, select a destination storage for the virtual machine files, and click Next. 7. Select the Disk Format in which to store the virtual disk: Thick Provision Lazy Zeroed Thick Provision Eager Zeroed Thin Provision (recommended) Deploying the Virtual Appliance on the vcenter server 27

28 Installation Eager Zero yields the best performance, but also takes the most time to initialize. Thick provisioning does not fill the drive unless eager zeroed is selected. If thick provisioning is selected, the storage capacity for the entire virtual disk is allocated on the data store at virtual disk create time. Thin provisioning means that the capacity on the data store is allocated to the virtual disk as required, up to the full size of the virtual disk. 8. Click Next. 9. In the Network Mapping window, specify what networks the deployed template use, and click Next. 10. In the Ready to Complete window, verify that the options are correct and click Finish, or click Back to change options. 11. When the deployment completes successfully, right-click the newly deployed Virtual Machine and select Template > Convert to Template. Converting the virtual machine to a template provides the ability to deploy multiple nodes. 12. When the conversion completes, continue to Deploying the virtual machine from a template. It is recommended that you rename the template to something intuitive, for example DPSearch. Deploying the vapp from a template It is recommended that you deploy the virtual machine from the template to provide the ability to add additional nodes as required. Procedure 1. Right-click the virtual machine and select Deploy Virtual Machine from this Template In the Name and Location window, specify a Name and an Inventory Location for this virtual machine, and click Next. 3. Select a Resource Pool within which to run the virtual machine and click Next. 4. In the Storage window, select a destination storage for the virtual machine files, and click Next. 5. In the Guest Customization window: Do not click the checkbox to Power on this virtual machine after creation. Select the Do not customize option. Click Next. 6. In the Ready to Complete window, verify that the options are correct and click Finish, or click Back to change options. A virtual machine clone is created from the Template. 7. When virtual machine clone is creation completes, right-click the computer and select Edit Settings Data Protection Search 1.1.x Installation and Administration Guide

29 Installation Add virtual disks on the vcenter Server 8. Continue to Customizing the Virtual Machine after deployment. If required, you can add virtual hard disks to provide storage for index data. Additional hard disks must be configured for all-in-one nodes and data nodes. Additional disks are not required for dedicated worker nodes. Procedure 1. Need a starting point. 2. Right-click the virtual machine and then select Properties. The virtual machine Properties window opens. 3. In the Properties window, on the Hardware tab, click Add. 4. In the Choose the type of device you wish to add list: a. Select Hard Disk. b. Click Next. 5. In the Select a disk window: a. Select Create a new virtual disk. b. Click Next. 6. In the Create a disk window, complete the following sections, and then click Next. Capacity (disk size) Disk Provisioning: Thin provision is recommended Location 7. In the Advanced Options window, click Next to accept the default settings. The settings on this page do not typically change. Configure the Virtual Machine on a vcenter Server You can configure the virtual machine after installation. Complete the virtual machine configuration immediately after installation. Changing the virtual machine settings later can make the virtual machine unstable. Procedure 1. In the Virtual Machine Properties window: configure the following settings: Memory CPU Disk 2. Add and configure additional disks as described in Adding virtual disks. Add hard drives for the index data. Deploying the vapp from a template 29

30 Installation 3. To open the console: a. Select the newly deployed virtual machine in the vcenter server. b. Select Power on from the list of commands. Deploying the vapp from VMware vcloud Director You can deploy the virtual machine by using VMware vcloud Director is available. The following figure illustrates the VMware vcloud Director. Figure 4 VMware vcloud Director Procedure 1. Log in to the VMware vcloud Director. 2. To begin the deployment, click Add vapp from OVF. 3. Browse to the OVF file. 4. Complete the wizard, accepting the defaults with the exception of the Computer Name. 5. In the Configure Networking section of the Wizard, change the Computer Name, and click Next. 6. If required, modify the following components in the Customize Hardware section. CPU: Number of virtual CPUs Cores per socket Number of sockets Memory Hard Disks 30 Data Protection Search 1.1.x Installation and Administration Guide

31 Installation Adding virtual disks with vcloud 7. Click Finish. It can take some time for the new vapp creation to complete. If required, configure new virtual hard disks. Additional hard disks must be configured for all in one nodes and data nodes. Additional disks are not required for dedicated worker nodes. Additional disks provide storage for index data. Procedure 1. Right-click the virtual machine and select Properties. 2. In the Hardware tab of the Virtual Machine Properties window, click Add in the Hard Disks section of the window. Data Protection Search installation planning and considerations provides details on recommended disk size. 3. Accept the defaults for the remaining fields, and click OK. Customizing the Virtual Machine in vcloud after deployment You can customize the virtual machine in vcloud after deployment. Before you begin Complete the virtual machine configuration immediately after deployment as changing the virtual machine settings later can make the virtual machine unstable. Procedure 1. In the Virtual Machine Properties window, configure Memory, CPU, and Disk. Data Protection Search node component requirements provides the specific recommendations for these settings. 2. In the Guest Customization tab, click to disable Allow local administrator password. If Allow local administrator password remains enabled, vcloud generates a random root password for the virtual machine. If Allow local administrator password is disabled, the default password, Linux remains. 3. Add and configure additional disks as described in Adding virtual disks with vcloud. Add additional hard drives to hold index data. 4. To open the console, click the newly deployed virtual machine in the vcenter server, and choose Power on from the list of commands. Initializing the Data Protection Search environment Data Protection Search configuration and general operation is handled through a webbased Administration console. However, some settings must be configured by using a menu system in the Linux terminal. There are two methods of configuration, basic (recommended default settings that are excluded from wizard) and advanced. Initializing the Data Protection Search environment 31

32 Installation Accepting the End User License Agreement (EULA), and configuring network settings must be completed before the rest of the options are displayed: Accepting the End User License Agreement Configuring network settings Configuring an LDAP server Providing LDAP Admin user and group accounts Specifying the criteria for the indexing subsystem Setting the local time and time zone Updating system passwords Set the Puppet role for future upgrades Procedure 1. Login to the Data Protection Search terminal: Username: dpsearch Password: dpsearch The default password is dpsearch. Change the password when possible for security reasons. 2. Type the following commands: su Type the root password: linux The default password is linux. Change the password when possible for security reasons. For versions earlier than Data Protection Search 1.1 SP3, enter cd download to change the directory with the install scripts. 3. Type the bash dp_install.sh command. The option to Show EULA opens. 4. To display the EULA, type To accept the terms of the EULA, type yes. 6. To exit the EULA, type q. 7. Type 2 to initialize the environment. The YaST2 menu opens to the YaST Control Center to configure network settings. Use the arrow keys, Tab, and Enter keys to browse and Alt-option to select items. 8. Tab to Network Devices and select Network settings from the list of devices. The Network Settings window displays details of the current Data Protection Search host: Device name and type IP address 32 Data Protection Search 1.1.x Installation and Administration Guide

33 Installation Bus ID 9. Type Alt -I to edit the eth0 device. 10. Set the following network options if you are using a static IP, and type Alt -N or tab to Next: IP address Subnet mask Hostname 11. Tab to Hostname/DNS and type Alt -S to set the following options: Hostname Domain name Name servers Domain search 12. Tab to Routing, or Alt -u to set the Default gateway. A default gateway is required. If one is not set, the deployment scripts display errors during firewall configuration. 13. To complete the network options and exit the tool, select OK, or Alt -o YaST2. If you change the hostname, a system restart is required for the change to take effect, type yes to restart. 14. Before continuing with installation on the Data Protection Search node, switch to the DNS server to configure the host lookup. Right-click the DNS/Active Directory/Open LDAP server, and select New Host: To add a record to resolve the hostname of the server to its IP address, type the node name in the Name field. Add the IP address defined in step 9. Ensure that nslookup returns the correct hostname/ip. For NetWorker, the host IP address must resolve to the same hostname defined for the NIC. Click Add Host. 15. When the Data Protection Search node restarts, complete the following: Log in with username: dpsearch Password: dpsearch Log in as root, su Type the root password: linux To verify that the Active Directory/Open LDAP server it is configured for the Data Protection Search node, use the ping command Type the bash dp_install.sh command again to restart the installation process Initializing the Data Protection Search environment 33

34 Installation The default password is dpsearch, and linux for the root user. Change both passwords when possible for security reasons. Basic configuration Both Advanced Configuration, and Basic Configuration options are available for Data Protection Search. Use the Advanced Configuration option if additional customization is required for the environment. To complete the recommended Basic Configuration, follow the instructions in this section. Procedure 1. Login to the Data Protection Search terminal: Username: dpsearch Password: dpsearch The default password is dpsearch. Change the password when possible for security reasons. 2. Type the following commands: su Type the root password: linux The default password is linux. Change the password when possible for security reasons. For versions earlier than Data Protection Search 1.1 SP3, enter cd download to change the directory with the install scripts. 3. Type the bash dp_install.sh command. The available setup options are shown in the following figure. 34 Data Protection Search 1.1.x Installation and Administration Guide

35 Installation Figure 5 Basic Configuration options 4. Type 3 to select Basic Configuration, and display the options available for Basic Configuration. 5. Type 1 to install the first node (Index Master node), and respond to the following prompts: Please enter the Elasticsearch Cluster name (default: DPSearchCluster): Provide a unique cluster name LDAP Type (for example, AD or openldap, default: AD): LDAP hostname (for example, ldap.domain.com): Provide the hostname of the AD or OpenLDAP server LDAP port (for example, 389): Provide the port for the selected LDAP server. For AD, the default port is 389 LDAP Base DN: (For example, dc=domain, dc=com): Provide the base domain for the LDAP server LDAP query username: Provide a username with rights to query the LDAP server (this user also administers CIS) LDAP query password: Provide the password for the username provided The user credentials that are provided are validated, and the second disc drive is partitioned and mounted. Data Protection Search Admin Group name (A valid group in AD or OpenLDAP that contains Data Protection Search Administrators, such as the Data Protection Search Admin Group). Provide the name of the existing LDAP security group, for example, DPSearch Admin Group. 6. When the Data Protection Search Admin Group is confirmed, you are prompted to Press any key to continue... The installation takes several minutes to complete. Basic configuration 35

36 Installation 7. Type 0 for Back to last menu. 8. Type6 for Change date and time: and select the options when prompted: Please select a continent, ocean, "coord", or "TZ". The available continents, oceans, coord, and TZ options are displayed. Please select a country whose clocks agree with yours. The date and time zone for the choices are displayed. Type Y when prompted, Is the above information OK?, or make the required changes. 9. Type 7 for Puppet Configuration: a. Type 1 for Configure as Puppet master. b. Type 0 for Back to last menu. c. Type 0 to exit the setup menu. 10. Restart, and change to the download directory after logging in to the terminal: dpsearch dpsearch /home/dpsearch/ 11. Run the following command to validate the installation: bash validate_dpsearch_install.sh -a 12. Verify that you can launch the Data Protection Search web application and log in. 13. Add additional Index Data/Worker nodes as described in the following sections: Installing Index Data components Installing additional Worker nodes Advanced configuration Both Advanced Configuration, and Basic Configuration options are available. Although Basic Configuration is recommended, use the Advanced Configuration option if additional customization is required for the environment. Procedure 1. Log in to the Data Protection Search terminal: Username: dpsearch Password: dpsearch The default password is dpsearch. It is recommended that you change the password for security reasons. 2. Type the following commands: su Type the root password: linux 36 Data Protection Search 1.1.x Installation and Administration Guide

37 Installation The default password is linux. Change the password when possible for security reasons. For versions earlier than Data Protection Search 1.1 SP3, enter cd download to change the directory with the install scripts. 3. Type the bash dp_install.sh command. 4. Type 2 to initialize the environment. To configure network settings, the YaST2 menu opens to the YaST Control Center. To browse, and select items, use the arrow, Tab, and Enter keys. 5. Tab to Network Devices and select Network settings from the list of devices. The Network Settings window displays details of the current Data Protection Search host: Device name and type IP address Bus ID 6. Type Alt -I to edit the eth0 device. 7. Set the following network options if you are using a static IP, and type Alt -N, or tab to Next: IP address Subnet mask Hostname 8. Tab to Hostname/DNS and type Alt -S to set the following options: Hostname Domain name Name servers Domain search 9. Tab to Routing, and set the Default gateway. A default gateway is required. If one is not set, the deployment scripts display errors during firewall configuration. 10. Select OK, and exit the YaST2 tool. If you change the hostname, a system restart is required for the change to take effect, type yes to restart. 11. Before continuing with installation on the Data Protection Search node, switch to the DNS server to configure the host lookup: Add a record to resolve the hostname of the server to its IP address Ensure that nslookup returns the correct hostname/ip For NetWorker, the host IP address must resolve to the same hostname defined for the NIC Advanced configuration 37

38 Installation 12. When the Data Protection Search node restarts, complete the following: Log in with username: dpsearch Password: dpsearch Log in as root, su Type the root password: linux Type the bash dp_install.sh command again to restart the installation process The default password is dpsearch, and linux for the root user. It is recommended that you change both passwords for security reasons. 13. Continue with the configuration process depending on the role you intend the Data Protection Search node to have: Single node (all in one) Follow the steps in Installing the Common Index Service (CIS). Follow the steps in Installing Worker components. Multi node Dedicated Index Master node (Do not configure more than one Index Master node). Follow the steps in Installing the Common Index Service (CIS). Dedicated Index Data node. Follow the steps in Installing Index Data components. Dedicated Worker node, perform either of the following: Follow the steps to install Worker Components (first Worker Node). Follow the steps in install Additional Worker Nodes (subsequent Worker Node). The CIS node must be installed as the first Data Protection Search node before any other nodes are installed. There can only be one Index Master node with these components. Installing the Common Index Service (CIS) CIS must be installed on one node in the environment, with Data Protection Search on an all in one node or on a separate node in a multi-node environment. The steps in this section are required only for Data Protection Search Advanced configuration. If the Advanced configuration option was selected from the dps_install.sh script, use the following procedure to continue with the CIS installation. The Basic configuration option provides default/recommended settings for the configuration options available in this section. 38 Data Protection Search 1.1.x Installation and Administration Guide

39 Installation Procedure 1. Type 4 to launch Advanced configuration, then type 1 to configure CIS. A submenu opens with the following options: Configure as Index Master Node (There can only be one CIS Master node) Configure as Index Data node (Many index data nodes can be added) Update LDAP Settings (used to update the LDAP query user password) 2. Respond to the following CIS installation prompts to configure the Elasticsearch nodes: Elasticsearch cluster Name Provide a unique cluster name Elasticsearch node name Accept the default Elasticsearch node heap size usually, choose half the physical memory of the node (usually this value is already the default) Number of index replicas The Number of index replicas (the default 1 is hard coded here) is used to determine how many replicas are made of each index. If you plan to install more than one index data node, you must set the replica number to >1, and lower than the (node number - 1). Change the number of replicas from the default of 0 in the Options section of the Admin UI. If there are not enough nodes to create the requested number of replicas, the status of the cluster and individual indexes remain yellow in the Data Protection Search Admin UI Dashboard. Number of shards Accept the default if you are not familiar with shard settings. The Elasticsearch website provides details on shards, and recommended settings. 3. Respond to the prompts to configure the LDAP settings: Select AD or OpenLDAP LDAP hostname name: Set to the name or IP of the OpenLDAP or Active Directory server (FQDN) LDAP Port: Port number (usually 389) LDAP Base DN: Base distinguished name of the domain or OU Data Protection Search uses when managing users and groups. For example: DC=domain, DC=com enables any users/groups in domain.com to be used OU=IT, DC=domain, DC=com enables any users/groups in the IT organizational unit of domain.com to be used LDAP Query Username: Account of a user that can query LDAP - Specify as username@domain.com (UPN for AD), or cn=username, dc=domain, dc=com (DN for OpenLDAP) This account is also the CIS Admin Advanced configuration 39

40 Installation LDAP Login Password: Password for the account The LDAP settings are validated. If the authentication fails, you are prompted to re-enter the information. 4. Press a key when prompted, Press Any key to continue... The Elasticsearch data storage location is set and the second disk drive for the virtual machine (dev/sdb) is: Partitioned Formatted Installing Index Data components Mounted to /mnt/elasticsearch_data Elasticsearch settings (yml) are updated Elasticsearch services are restarted 5. Type 0 to return to the main menu when the new window is displayed. You can configure Index Data components by using the DPSearch & CIS Deployment script. Before you begin The Index Master node must already have been installed and configured. The Index Master automatically includes the Index Data components. It is not necessary to perform these steps on the Index Master node. From the DPSearch & CIS Deployment script, you can install additional index data nodes. Procedure 1. To configure CIS, type 1 to launch Advanced configuration, and then type Type 2 in the CIS Install menu: Existing DP machine name: Enter the name of the Index Master node (the first node installed) Elasticsearch Cluster name: Accept the default (unless changed for first node) Elasticsearch node name: Accept the default (name of this node) Elasticsearch node heap size: See nodes for the initial CIS install Number of Index Replicas: Same as the Index Master node Number of Index Shards: Same as the Index Master node 3. Repeat these steps to install additional index data nodes. 4. Verify that the index data nodes are successfully added in CIS nodes in the System section of the Admin UI. The nodes are listed. 40 Data Protection Search 1.1.x Installation and Administration Guide

41 Installation Installing Worker components Before you begin The Index Master components must already have been installed and configured on either this node, or on another node. Procedure 1. Type 4 to begin Advanced configuration, then type 1 to configure Data Protection Search. 2. Configure the initial Data Protection Search node: Install directory: Accept the default although it can be modified. CIS URL: If CIS was installed on this node (Index Master), accept the default. If CIS was installed on another node, type the URL of that node. Use the same format as the default, but change the node name. CIS Admin: Type the username and password for the CIS Admin configured in Installing the Common Index Service (CIS) Data Protection Search Application name: Accept the default Worker node: Accept the default (yes), unless you are creating customized Worker nodes Web Services node: Leave as default (yes), unless you are creating customized Workers Data Protection Search Admin Group name: The name of a security group in LDAP While the installation is validated, the following occurs: The CIS Admin is logged in to CIS The DPSearch Admin group is validated in LDAP You are prompted to Press any key to Continue Follow the prompt to Press any key to Continue... The script will: Log in to CIS Create the DPSearch application Add the current node Setup system configuration Create the system index Initialize the system index mappings Configure the system index Create system activities Configure the worker Configure LDAP The Data Protection Search & CIS Deployment script opens. 4. Type 0 to open to the main menu. 5. Type 6 to change the date & time. Ensure that the date and time zone are the same as the backup servers. Installing Worker components 41

42 Installation 6. Exit the installation script. Results Installing additional Worker nodes You can now log in to the DPSearch Admin UI. Use the following procedure to install additional Worker nodes. Before you begin The initial Worker node and the Index Master node must be installed and configured. Procedure 1. Type 4 to start Advanced Configuration. 2. Type 2 to Configure Data Protection Search. 3. Type 2 to Configure an additional Data Protection Search Node. Consider the following: Table 10 Prompt descriptions Prompt Install directory CIS URL CIS Admin DPSearch Application name Worker node Web Services node Description Can be modified. Modifying the install directory path is not recommended. Use the CIS URL for the Index Master node. Use the full URL as in the default. Use the same username and password that is used for the Index Master node. Use the default DPSearch Application name unless the name was changed on the initial node. Use the default option unless you are creating customized Worker nodes. Use the default unless you are creating customized Worker nodes. 4. To install additional Worker nodes, repeat these steps. Installing a self-signed or trusted certificate The NGINX web server that is provided with Data Protection Search is installed with a self-signed certificate, not a trusted public key certificate. The certificate is used for secure http access (https) to the web user interfaces, Admin, and Search REST APIs, and the Common Indexing System (CIS) REST API. The certificate includes secure communications between the components. When a self-signed certificate is active, users connecting to the web-based Admin and Search interfaces are warned that they are connecting to an untrusted connection. For most web-browsers, this warning can be suppressed after it is initially displayed. To install either a self-signed, or trusted certificate for the Data Protection Search NGINX web server, perform the following steps: 42 Data Protection Search 1.1.x Installation and Administration Guide

43 Installation Procedure 1. Connect to the Data Protection Search node as root, and use the default password linux. 2. Copy the existing certificate and private key files to a backup location: cp /etc/nginx/dpsearch.cert /BACKUP LOCATION cp /etc/nginx/dpsearch.key /BACKUP LOCATION 3. (Optional) Generate a new private key: openssl genrsa -out dpsearch.key Complete either of the following: Type the following command to create a self-signed certificate by using the existing or newly generated private key file: openssl req -new -x509 -key dpsearch.key -out dpsearch.cert - days 1095 Respond to the prompts. Type the following command to generate a certificate request (csr) file by using either the existing or newly generated private key file: openssl req -new -key dpsearch.key -out dpsearch.csr a. Respond to the prompts. b. Send the dpsearch.csr file to the certificate authority. c. Rename the returned certificate file to dpsearch.cert. 5. Stop the NGINX service: service nginx stop 6. Copy the new certificate to the /etc/nginx directory. Optionally, copy the new private key to the /etc/nginx directory: cp dpsearch.cert /etc/nginx/ cp dpesearch.key /etc/nginx/ 7. Verify that the files have the correct permissions: chmod 644 /etc/nginx/dpsearch.cert chmod 644 /etc/nginx/dpsearch.key 8. Start the NGINX service: service nginx start Update LDAP configuration During deployment, the configuration of an LDAP server must be specified. At a later date, some of the specified settings must be updated. Change the password of the LDAP query account, or the account name itself. There are two ways to update LDAP configuration as described in the following tasks. Update LDAP configuration 43

44 Installation Updating LDAP configuration in the Data Protection Search Admin UI During deployment, the configuration of an LDAP server must be specified. At a later date, some of the specified settings must be updated. Change the password of the LDAP query account, or the account name itself. Procedure 1. Log in to the Data Protection Search Administration UI by selecting the following: You will have to log in to Data Protection Search Administration UI each time it is opened, or after an inactivity time out (1 hour by default). 2. Select Administration > Options > LDAP Options. 3. In the Host field, type the host name of the LDAP server. 4. In the Port field, type the port number that is used by the external authentication authority: For LDAP, the default port number is 389. For SSL, you can use port In the Base DN field, type the scope of the users and groups that are considered within the LDAP server. For example: DC=example, DC=com The Base DN determines the structure of the LDAP server where the search filter is applied. This is usually similar to the domain name over which the LDAP server has authority. 6. In the Admin User field: a. Type a user account that has full read access to the LDAP directory, in the following format: user@domain. For example, administrator@ldap.example.com b. Ensure that the username is one of the following: Common name address Entry distinguished name c. Ensure that the user has read access to the directory. d. To include notification, define the address for the account. 44 Data Protection Search 1.1.x Installation and Administration Guide

45 Installation Only admin accounts with defined addresses can receive notifications. The default admin user is not configured with an address and cannot receive notifications. 7. In the Password field, type the password of the user account that you specified in the Admin User field. 8. In the SSL field, select either of the following options: To not apply secure connection settings, select Off. This option is the default setting. To connect to an external authentication server using LDAPS, select On. The Verify Certificates field appears. 9. To verify certificates: a. In the Verify Certificates field, select On. b. Copy the PEM files to the appropriate directory: For Data Protection Search 1.1 SP3: /etc/pki/trust/anchors/ For Data Protection Search 1.1, 1.1 SP1, and 1.1 SP2: /usr/share/ca-certificates/ At the command prompt, type the following command: update-ca-certificates 10. Click Validate. Updating LDAP configuration in the Data Protection Search Admin installation script If the LDAP query user password has been reset, it may not be possible to log in to the Data Protection Search Admin UI. The LDAP user is required to authenticate the login user against the directory service. If required, modify the LDAP configuration in a terminal session with an SSH tool such as PuTTy for the Data Protection Search Index Master virtual machine. Procedure 1. Login as root (default password is linux). 2. Change to the install directory. 3. For versions earlier than Data Protection Search 1.1 SP3, enter cd download to change the directory with the install scripts. Run the dp_install.sh script: bash dp_install.sh 4. Select option 4, Advanced Configuration. Updating LDAP configuration in the Data Protection Search Admin installation script 45

46 Installation 5. Select option 1, Configure CIS. 6. Select option 3, Update LDAP settings. 7. To update the settings, follow the prompts. Updating the Data Protection Search Admin Group If the LDAP domain changes, it might be necessary to change the Data Protection Search Admin Group. Procedure 1. Open a terminal session for any Data Protection Search Worker node virtual machine using an SSH tool such as PuTTy. Log in as root with the default password linux. 2. Change to the /bin subdirectory of the installation directory, The default is /usr/local/dpsearch/bin. 3. Change the Data Protection Search Admin Group in the admin_cn field of ldap_settings. 4. Save the file, and exit the editor. 5. Run the following command: ruby config_system.rb -o set_group -f config.json Upgrading to the current release of Data Protection Search An upgrade.zip package (dpsearch-upgrade-build_number.zip) is available to download from the same location as the binaries. Procedure 1. Download the.zip file and unzip it by using an appropriate program. There are two directories: Puppet Puppet is an open source configuration management tool that provides the ability to deploy the update files from a single node to all nodes in a multinode, Data Protection Search Cluster environment. Puppet must be installed on each node in the cluster only when upgrading from Data Protection Search 1.0. Upgrade 2. Using a secure FTP client such as WinSCP: Copy both the /upgrade and /puppet directories to the Data Protection Search Index Master Node. Copy the puppet directory to each node in the cluster. 3. In a Console window (SSH), browse to the /puppet subdirectory of the directory to which you copied the files, and log in as root user: su Password 46 Data Protection Search 1.1.x Installation and Administration Guide

47 Installation 4. Run the puppet install script: bash config_puppet.sh 5. In the Data Protection Search Index Master Node, type y to confirm that it is the puppet master: would you like to configure this node to work as puppet master? (y)es or (n)o: The installation completes. If the deployment is an all-in-one environment, configure it as puppet master as described in step 5, and skip to step 7 directly to upgrade this all-in-one node. 6. For the remaining nodes in the cluster: a. Browse to the /puppet subdirectory of the directory to which you copied the files, and log in as root user: su Password b. Run the bash config_puppet.sh script. c. Type n for the following prompt, to specify that it is a puppet agent: would you like to configure this node to work as puppet master? (y)es or (n)o: d. When prompted, provide the location for the puppet master on the Data Protection Search Index Master Node. Steps 4 6 are only used to configure the Puppet environment. It is necessary to complete steps 4 6 when upgrading from Data Protection Search1.0 to Data Protection Search 1.1. Steps 4 6 can be skipped for all subsequent upgrades. Also, to change the puppet role from a single node, manually delete the /etc/ puppet/puppet.conf file and re-run the bash config_puppet.sh script. For a new Data Protection Search 1.1 installation, it is not required to configure the puppet master, and steps 4 6 are unnecessary. The Data Protection Search Admin can specify the puppet master and puppet agent roles from the installation menu. Upgrading to the current release of Data Protection Search 47

48 Installation Figure 6 Configuration script options 7. From the Data Protection Search Index Master Node, browse to the directory in which you copied the files (parent folder of the /puppet subdirectory). To provide access permissions for the puppet files, run the following command: chmod 777 -R upgrade/ chmod 777 -R puppet/ 8. On the Data Protection Search Index Master Node, browse to the /upgrade subdirectory of the directory to which you copied the files, and run the upgrade script: bash update.sh -o deploy The upgrade is applied to all nodes in the cluster. There are many files so the upgrade can take some time to complete. 9. (Optional) Run a report to view the status of the upgrade, and to verify that the upgrade completed successfully for all nodes in the cluster: bash update.sh -o report The report (report_date_unique identifier.csv) is available in the /download/ upgrade/report directory. 10. Use a secure FTP client such as WinSCP or PuTTy to copy the report and view it in an application such as Microsoft Excel. 11. View the update logs from the/root/.dpsearch_update/log/ directory: ls /root/.dpsearch_update/log -l The log files are useful if the upgrade fails for any nodes in the cluster. Results When you log in to Data Protection Search from the web browser, you see the login dialog box, and the dashboard for the current release of the product. In the Workers section of the Admin dashboard, the Workers are listed and the new build number is displayed in the respective details. 48 Data Protection Search 1.1.x Installation and Administration Guide

49 CHAPTER 3 Administration This section includes the following topics: Data Protection Search Administration Web Application Logging in to the Data Protection Search Admin user interface...50 Data Protection Search Admin UI home...51 Data Protection Search dashboard Administration 49

50 Administration Data Protection Search Administration Web Application The Data Protection Admin user interface provides the ability to administer, configure, and customize Data Protection Search. When the Data Protection Search virtual appliance is deployed and the web server is configured and running, you can access the Data Protection Search Admin web app hosted by any Worker node. Logging in to the Data Protection Search Admin user interface Before you begin To log in to the Admin UI initially, you must be a member of the Data Protection Search Admin group (LDAP user) configured during installation. Later, additional Index Admins can be created to log in and access the Admin UI. Cookies must be enabled in the browser. After completing the Data Protection Search virtual appliance deployment, log in to the Admin UI. Figure 7 Admin UI login Active Directory and OpenLDAP are both configurable in /usr/local/ dpsearch/etc/ldap.conf, by default: The following are examples of supported Active Directory username formats: SamAccountName (administrator) User Principal Name (administrator@domain.com) DistinguishedName (cn=administrator, cn=users, dc=domain, dc=com) Windows NT account (domain\administrator) Mail (administrator@domain.com) cn (administrator) 50 Data Protection Search 1.1.x Installation and Administration Guide

51 Administration The following are examples of supported OpenLDAP username formats: cn (administrator) Mail entrydn (cn=administrator, cn=users, dc=domain, dc=com) Procedure 1. Type the username in the User name field. 2. Type the password in the Password field. 3. Click Login. At initial login, all options of the Data Protection Search Dashboard are available, as the DPSearch Admin user is also an Index Admin. Log in to the Data Protection Search UI each time it is opened, or after an inactivity time out (1 hour by default). The default inactivity timeout can be modified in the Options section of the Admin UI. Data Protection Search Admin UI home The Data Protection Search Admin UI is customized based on role. Figure 8 Admin UI dashboard The following table lists and describes the sections of the Data Protection Search Admin UI. Table 11 Data Protection Search Admin UI Admin web UI tab Description Admin permission visibility DPSearch Admin Index Admin Dashboard View a summary of the health/status for the various Data Protection Search Components and drill down for more information. Yes, with the exception of Indexes and Scheduled Collections. Yes, can view Indexes, Scheduled Collections, and Notifications. Sources Add, update, or remove Avamar and NetWorker servers. Yes No Data Protection Search Admin UI home 51

52 Administration Table 11 Data Protection Search Admin UI (continued) Admin web UI tab Description Admin permission visibility DPSearch Admin Index Admin Roles Add, update, or remove Index Admins. DP Search Admins are listed, but cannot be modified. Instead, they are managed with the LDAP solution. Yes No Indexes Add, update, or remove indexes. Manage permissions for each Search Admin No Yes Collections Schedule collections for Avamar and NetWorker backup servers. No Yes Systems Monitor Data Protection Search Worker and Index nodes. Yes No Jobs Lists running and completed activities and jobs. Details include type, status, duration, and more. Yes Yes Options Modify or enable the following: Yes No Search Options, the number of Search hits to display Session Options (timeout) LDAP Options notification Options (On or Off) Index Options (Including file types to exclude from Full Content Index, Number of replicas, and Apply replica settings to existing indexes) Help Access the Data Protection Search online help. Yes Yes Data Protection Search dashboard After logging in to Data Protection Search Admin UI, the full dashboard opens displaying a summary of the health/status of the components that make up Data Protection Search environment. Each section is color coded, and there is a section to display the number of healthy (green), warning (yellow), and error (red) items for each component. To view a detailed list and additional information, click the expand/collapse arrow to the right of each dashboard component name. For Scheduled Collections and Notifications, click More... to view a detailed list and additional information. To force a data refresh, there is a refresh icon for each component. 52 Data Protection Search 1.1.x Installation and Administration Guide

53 Administration For Servers, a Last Updated Time field provides a timestamp for unresponsive servers. If the Last Updated Time value remains the same after approximately 1 hour, it is listed in a red status bar, and a message similar to the following is displayed: One or more Servers are unresponsive For Workers, a Last Heartbeat Time field provides a timestamp for unresponsive workers. If the Last Updated Time value remains the same after approximately 10 minutes, it is listed in a red status bar, and a message similar to the following is displayed: One or more Worker services are unresponsive Figure 9 Data Protection Search Admin dashboard The following table lists the dashboard components and visibility based on Admin permissions. Table 12 DP Search dashboard Health/ Status for component Servers Description Lists up to 5 configured backup servers with information on platform, version, and status. From the bottom of the expanded list of servers, click More... to open the Sources section of the Admin UI to view additional information. The last time the status of each backup server was updated is displayed. To refresh the server status, click the refresh icon. Admin permissions visibility DP Search Admin Index Admin Both DP Search and Index Admin Yes No Yes Data Protection Search dashboard 53

54 Administration Table 12 DP Search dashboard (continued) Health/ Status for component Workers Clusters Indexes Scheduled Collections Notification Description Lists up to 5 configured Data Protection Search Worker nodes and their status. To see a last updated time, and view unresponsive worker services, click More... at the bottom of the list of expanded Worker nodes. The System section of the UI opens. Lists up to 5 configured Elasticsearch Cluster status. More...To open the System section of the DPSearch UI, click More...at the bottom of the list. Lists up to 5 configured indexes and their status. At initial login, no indexes have been created so the list is empty. More...To open the Indexes section of the DPSearch UI, click More... at the bottom of the list. Lists up to 5 upcoming scheduled collections. To view the full list of configured collections, click More... at the bottom of the list. Displays up to 5 recent notifications. To view and manage notifications, click More... Configure notifications in Notification Options in the Options section of the Admin UI. You can open System Notifications, mark them as read, unread, or view all notifications. Admin permissions visibility DP Search Admin Index Admin Both DP Search and Index Admin Yes No Yes Yes No Yes No Yes Yes No Yes Yes Yes Yes Yes 54 Data Protection Search 1.1.x Installation and Administration Guide

55 CHAPTER 4 Sources This section includes the following topics: Sources...56 Adding an Avamar server to Data Protection Search Default Avamar server limit Adding a NetWorker server to Data Protection Search...58 Connection Limitations considerations...59 Updating an Avamar or NetWorker server Removing a server from Data Protection Search...61 Registering agents manually Sources 55

56 Sources Sources Add, update, and remove Avamar and NetWorker servers in the Sources section of the Data Protection Search Admin UI. At initial login, the Data Protection Search Admin is prompted to add sources. A check mark in the server product icon (Avamar or NetWorker) confirms that there is a connection to the respective server. If Data Protection Search is unable to connect to the server, a warning icon is displayed. You can also use Find by name... to find a server from the list. Figure 10 Sources Adding an Avamar server to Data Protection Search You can add Avamar servers to Data Protection Search. Procedure 1. In the Sources section of the UI, click the add icon ( ). The Add Source window opens. 2. In the Add Source window, complete the required fields as listed in the following table. 56 Data Protection Search 1.1.x Installation and Administration Guide

57 Sources Table 13 Required felds Required fields Name Platform Description Type the source hostname or IP of the Avamar server. Select Avamar from the drop down list. Port Default is port User ID Can be an existing Administrator user, an Avamar Management Console user, or a Data Protection Search user. You can choose an existing Avamar Admin user for that server or create your own in the Avamar Administrator. Password Confirm Password Timezone Connection Limitation Provide the password for the corresponding user as described in User ID. Re-enter the password. Choose a timezone for the current Avamar server from the drop down list. Click the check box to enable a Connection Limitation. Set a limitation, or leave the fields (Indexing, Restore and Download) blank for unlimited connections. The Connection Limitation is disabled by default. Enable blackout window Enable only for full content indexing Click the check box to enable and select the specified time range each day that no collection jobs can run for that backup server. Click the check box to restrict full-content collection activities during the blackout window. If checked, metadata only collections can run at any time. 3. Click Add. The new server is now available in the list of servers in Sources. When a server is added, a system job is added to register the Avamar client on each Data Protection Search Worker node to the new server. This can be monitored in Jobs by clicking the System Jobs checkbox. When the Avamar domain/clients is missing, adding an Avamar server as a source in Data Protection Search is not possible. Adding an Avamar server to Data Protection Search 57

58 Sources Default Avamar server limit Data Protection Search can index multiple Avamar servers simultaneously. Data Protection Search registers multiple Avamar File System clients on each Data Protection Search node. Each client is configured to use a different Transmission Control Protocol (TCP) port to communicate with the server, beginning at By default, for Avamar 7.1 and later, the Avamar server only exposes ports through the firewall ranging from to Once a port above is used, the server is unable to connect to the client. If more than 10 Avamar servers are configured in Data Protection Search, the server is unable to connect to the client. Removing and re-adding a server does not work. Adding an Avamar server to Data Protection Search succeeds. However, full content indexing or download operations do not complete. Browsing Data Protection Search nodes in the Avamar Administration GUI result in a browse timeout. Client details display a page port of or higher. To workaround this issue, the port range must be increased for additional servers. The /etc/firewall.default file on the Avamar server must be edited. The exec_rule -A OUTPUT -p tcp -dport 28001: j ACCEPT line must be modified to include a wider range. For example, exec_rule -A OUTPUT -p tcp -dport 28001: j ACCEPT. Once a port range is expanded on the Avamar server, existing Data Protection Search operations may not complete. If the operation does not complete, cancel and then rerun the operation. Adding a NetWorker server to Data Protection Search Before you begin 1. Register any Data Protection Search nodes as clients on the NetWorker server. 2. Specify a nominal file, or turn off scheduling (scheduling is not required to back up files). 3. For the User group: The root account on all Data Protection Search nodes must have the correct privileges. For example, user=root, host=dpsearchnode1.domain.com. For Option 1, add to the existing Application Administrator's or Operator's group. For Option 2, create Data Protection Search Admin's User Group. Add necessary root accounts. Provide the necessary privileges. The root accounts must have at least the same privileges as the Operator's group. 58 Data Protection Search 1.1.x Installation and Administration Guide

59 Sources Procedure 1. In the Sources section of the UI, click Add. The Add Source window opens. 2. In the Add Source window, complete the required fields as listed in the following table. Table 14 Required fields Required fields Name Platform Time zone Connection Limitation Description Type the source hostname or IP of the NetWorker server. Select Avamar from the drop-down list. Choose a time zone for the current NetWorker server from the drop-down list. To enable a Connection Limitation, click the checkbox. Set a limitation, or leave the fields (Indexing, Restore, and Download) blank for unlimited connections. The Connection Limitation is disabled by default. Enable blackout window Enable only for full content indexing To enable, and select the specified time range that daily collection jobs do not run, click the checkbox. To restrict full content collection activities during the blackout window, click the checkbox. If restricted, metadata only collections run. Configure all Data Protection Search nodes as a client of the NetWorker server before adding it. 3. Click Add. The server is now listed in Sources. When a server is added, a system job is added for each Data Protection Search Worker node, and the NetWorker registration process completes. The registration can be monitored in Jobs by clicking the System Jobs checkbox. Data Protection Search does not recognize a change in retention that is made on the backup server for 30 days. Connection Limitations considerations Connection Limitations provide the ability to limit to the number of concurrent connections available for Data Protection Search to the backup server. Limiting the number of concurrent connections to the server leaves resources for necessary operations such as scheduled back ups. Set limitations in the following three fields: Connection Limitations considerations 59

60 Sources Indexing Restore Download Leaving Connection Limitations fields blank permits unlimited connections between Data Protection Search and the backup server. Indexing Indexing means connections that are made to index the metadata or full content of files, as part of a collection activity. If you have multiple Worker nodes, one collection activity can create many connections. Set a connection value to control the Worker nodes. Restore Restore means restore operations that are manually triggered from the Search application, by Search Admins. Download Download means download operations that are manually triggered from the search app. Downloads only include one file at a time, so that may be a lower priority to limit, as compared to indexing, for example, which may generate 100s of requests. Impact of Connection Limitations Limiting indexing connections results in slower Data Protection Search indexing. Limiting restore or download connections results in longer response times for restore or download requests for Search Admins. Updating an Avamar or NetWorker server You can update the configuration of an Avamar or NetWorker server that was previously added to Data Protection Search. Click the Avamar or NetWorker server, and click the edit icon ( window, perform one of the following: For an Avamar server, complete the following fields: ). In the Update Name Port User ID Password Time zone Connection Limitation Blackout window (the update takes effect at the next runtime) You can choose an existing Avamar Admin user for the server, or create a user in the Avamar Administrator. For a NetWorker server, the following fields can be updated: Time zone 60 Data Protection Search 1.1.x Installation and Administration Guide

61 Sources Connection Limitation Blackout window Removing a server from Data Protection Search Use the following procedure to remove Avamar and NetWorker servers from Data Protection Search. When a backup server is removed, any item that is indexed from that server remains in the index until the backups expire. However, it is not possible to download or restore these items. Indexed items cannot be restored, even if the backup server is re-added. The re-added server has a different internal identifier. It is recommended that associated indexes are removed for deleted servers. Procedure 1. Select the server from the list and click the Delete server icon ( ). 2. A message similar to the following is displayed: Selected item(s) will be permanently deleted. Please type DELETE to confirm. If there are outstanding collection activities that are defined, a server cannot be removed. Delete the collections first. Also, a source cannot be removed If registering jobs running on a source, it cannot be removed. For example, when the server status is initializing, or after manually creating registering jobs. 3. Type DELETE, and click Confirm. Results The server is no longer listed. Registering agents manually If a source failed to initialize, or the message, Some agents failed to register displays, some of the agents did not register correctly and might not work. Incomplete configuration caused by network connectivity, firewall issues, and so on, cause registration failure. If required, you can manually trigger registering jobs for a server to correct the problem. Procedure 1. Select the server from the list, and click the Register icon. A message similar to the following is displayed: Successfully created jobs to register backup server agents. The status of the jobs can be monitored on the Jobs page. 2. Click OK to close the message. Register If registering jobs are running on the server, the Register icon is disabled. Removing a server from Data Protection Search 61

62 Sources 3. Monitor the registering job status in Jobs > View system jobs, or refresh the Sources list. 62 Data Protection Search 1.1.x Installation and Administration Guide

63 CHAPTER 5 Roles This section includes the following topics: Roles UI...64 Data Protection Search Admin role Roles 63

64 Roles Roles UI When you configure the Data Protection Search virtual appliance, configure a DPSearch Admin Group. The Admin Group and its associated users (created in Active Directory, or Open LDAP) are listed in the Roles section of the UI. Roles is also where you can add, remove, and assign permissions to the Index Admin Groups. The DPSearch Admin Group is the default Index Admin. Members of the DPSearch Admin Group are listed and cannot be edited directly. DPSearch Admin users are added, and deleted in Active Directory, or Open LDAP. Data Protection Search Admin role Figure 11 Admins Roles The DPSearch Admin role provides the ability to: Manage backup servers Manage roles Monitor system Monitor system jobs Set options Receive notifications 64 Data Protection Search 1.1.x Installation and Administration Guide

65 Roles Index Admin role Figure 12 Roles Index Admins Adding Index Admins Index Admins have permissions to: Create metadata only indexing collection activities (default) Create metadata only and full content indexing collection activities (must specifically enable full content indexing capability) Create and maintain indexes Monitor index jobs Receive index jobs related to notifications Procedure 1. In the Indexes > Index Admin window, click the add icon: 2. In the Select User window: a. Type a username or a substring. b. Display the list of Active Directory, or Open LDAP users, by clicking Find. c. Restrict the search to Users only, Groups only, or both (the default). If you click find before specifying a substring, the entire directory is returned, which can be slow. For example, to search for all users that contain Admin, type Admin, and click Find. 3. Select the user or group to add and enable Metadata Index only (default) or Metadata and Full content Index. Adding Index Admins 65

66 Roles Full content index searches can take longer than Metadata index searches and put a substantial strain on the backup server and backup storage performance. Removing an Index Admin Use the following procedure to remove Index Admins from Data Protection Search. Before you begin You must have full Data Protection Search Admin privileges to remove Index Admin users or groups from Data Protection Search. Procedure 1. In the Indexes > Index Admin window, click the Delete icon ( ). 2. To remove users or groups, select the checkboxes, and click Remove. Results The users or groups are no longer listed. Data Protection Search Admin role If required, Data Protection Search provides the ability to create multiple indexes, and to specify the users and/or groups with access those indexes. These users/groups are referred to as Search Admins. When a Search Admin logs in to the Search UI, they can search only those indexes to which they have access. The following table lists the DPSearch Admin roles. Table 15 Admin roles Search Admin role Index Admin - All access Index Admin - Read only Description No restrictions are applied. Cannot view inline or full content preview for search hits, download files locally, or restore files to an alternate location. The Data Protection Search Admin Group is the default Index Admin. Members of the Data Protection Search Admin Group are listed and cannot be edited directly. DPAdmin users are added and modified in any LDAP based directory service, such as Active Directory. 66 Data Protection Search 1.1.x Installation and Administration Guide

67 CHAPTER 6 Indexes This section includes the following topics: Indexes view Adding an index...68 Editing an index...69 Removing an index Indexes 67

68 Indexes Indexes view Indexes hold the indexed metadata and/or content extracted from backup files. Data Protection Search provides the ability to create multiple indexes, and to specify the particular users and/or groups able to access those indexes. These users/groups are referred to as Search Admins. When a Search Admin logs in to the Search UI, they can search only those indexes to which they have access. All configured indexes are listed in the Indexes section of the UI. For each index, the following information is displayed: The name and description of the index The size and number of items in the index Information that is provided by Elasticsearch: Adding an index The number of items in the index is not exact, since additional records for backups/save sets are also stored in the index The date/time the index was created, and the last modified date/time Use the following procedure to add an index. Procedure 1. In the Indexes section of the UI, click the add icon ( ). 2. In the New Index window, complete the following fields: Index name Description Analyzer User/group When an index is created, the user that logged in is added as the default Search Admin for that index, and is assigned an Admin All Access role. If required, this user can be removed, or the role can be changed. 3. Additional or replaced users and/or groups can be added. To add a User/group for the index, click the add icon in the Users/Groups section of the window. To add a User/group for the index: a. In the Select user window, type the name of the user or group to add, and click Find. b. To select the user/group in search results, click the row, and then click OK. The user or group is now listed in the Users/Groups section of the window, and is assigned an Admin Read only role by default. c. To change from Admin Read only for the user or group, click Admin All access in the Roles column. 4. Select an Analyzer from the list. The Analyzer options include, Standard (recommended), Simple, Whitespace, and Language. 5. Click Finish. 68 Data Protection Search 1.1.x Installation and Administration Guide

69 Indexes Editing an index Results The new index is listed. Use the following procedure to edit existing indexes from the Manage Indexes window. Procedure Removing an index 1. To modify an index, select it, and click the edit icon ( ). The following index details are displayed: Index name Index description Add, or change the Index description. Analyzer Not editable. User/Group You can add to, or remove users or groups from the index. Apply or change the following permissions for a specified User/group: Admin: Read only Admin: All access 2. Make required changes, and click Update. Use the following procedure to remove an index. Procedure 1. To remove an index, select it, and click the delete icon ( ). 2. When prompted, type DELETE, and click Confirm. A message similar to the following is displayed: If you delete this index, its contents will be permanently lost. Type DELETE to confirm Results The index is no longer listed. Editing an index 69

70 Indexes 70 Data Protection Search 1.1.x Installation and Administration Guide

71 CHAPTER 7 Collections This section includes the following topics: Collections activities Adding a collection activity...72 Editing a collection activity Collections 71

72 Collections Collections activities The Collections section of the Admin UI provides the ability to create and configure collection activities for Data Protection Search. These collection activities are used to identify the backup clients to be indexed. The rules that define how indexing is applied, for example, the time and duration of the indexing. The following table lists the available Collections activity actions. Table 16 Available actions for Collections activities Action Add Edit Remove Run now Enable Disable Description Add a Collection activity. Modify a collection activity. Removes one or more existing Collection activities by selecting the activities and clicking Remove from the list of actions. Force one or more scheduled collection activities to run now (rather than waiting for the next scheduled time). Enable one or more collection activities. Activities that are enabled run based on the schedule, and can also be forced to run now. Disable one or more collection activities. Activities that are disabled do not start, regardless of the schedule. Adding a collection activity You can show or hide the taskbar, and filter the activities by Name and Status. To modify the filter for activities, click the icons to filter, select all, and refresh, or click Reset filter. Viewing collection activities To view details of all jobs that have completed or are currently running, click the activity name to view a summary of the configuration settings for that activity. In the summary, click Details in the Jobs row. To open the Manage Collection Activity wizard and modify the activity, click the edit icon ( ) from the details. The edit icon in the taskbar is enabled only when an item is selected. This section provides the ability to select Avamar and NetWorker backup clients to index. A wizard provides the ability to add and modify Collection activities in Data Protection Search. Procedure 1. In the Collections window, click the add icon ( ). The Activity Information section of the wizard opens. 2. Complete the following fields in the New Collection Activity window: Name (required field) 72 Data Protection Search 1.1.x Installation and Administration Guide

73 Collections Description (optional field) Index (Select an index from the drop-down list) Specify information to index: Document Metadata only (recommended) Full content, including metadata and text Full content indexing can be very time and resource consuming, and negatively impact Avamar and NetWorker backup server performance. It is recommended that metadata only indexing is the primary approach, with full Content indexing enabled only as needed. To obtain more information on selecting full content indexing, click impact in the disclaimer. Accept the disclaimer before you can continue if full content indexing is selected. To obtain an estimate of the impact for the environment, click Calculate: a. Select full content indexing. b. Click Save & Next. Calculate can take a very long time to query Avamar and/or NetWorker for the estimate, and provides a high estimate, or worst case. The Sources section of the wizard opens. 3. To display the list of backup servers, click the add icon ( ) from the taskbar in Sources. 4. To display the list of available clients, select one, or more Avamar or NetWorker backup servers. You can search or filter on the backup server or Clients list. 5. Select one or more backup clients from the list and click Add. If necessary, the Refresh button can be used to update the list of clients from the backup server. 6. Click OK when the required clients are added. A single collection can have clients from multiple servers, and from both Avamar and NetWorker servers. 7. Click Save & Next. 8. In the Scope section of the wizard: The Activity index type is listed Content filter type options are: Index all documents (*.*) (default) Index only the following documents: Content filter Adding a collection activity 73

74 Collections Excluded file types Included file types (separated by a semi-colon) Modified date (Any Date, Specific Date, Between, Before, After) 9. Click Save & Next. 10. In the Schedule section, specify when to index the selected content (ASAP is the default): The schedule section is used to define the time, duration, and the collections recurrence schedule for indexing. Configure collection activities to recur daily or weekly to match the backup schedule. Matching the backup schedule ensures that new backups are processed for indexing when they complete. Therefore, the indexing window should be scheduled to occur after the backup window. By default, the schedule is set up to run a single time, when possible (ASAP). Recurrence pattern, select from: Once Daily Every n days on: Weekly Every n weeks on: Select the day of the week Start date Start time Duration, choose from the following: Run until completion Not more than n hours Collection activity fields n is an integer between 1 and 8. Range of occurrence: No end date End by, select a date for the schedule to end A list of Collection Activities is displayed. 11. Click Save and Next. 12. The wizard completes with an Activity Summary, and when you click Finish, the activity status changes to Pending. If required, click Back to modify the Collection Activity. The following table lists the information available for collections activities in the Collections section of the Admin UI. 74 Data Protection Search 1.1.x Installation and Administration Guide

75 Collections Table 17 Collection activity fields Field Name Description Index Schedule Next run Status Description Click the activity name to view a full summary of the activity details. Metadata only or full-content indexed and associated details. Lists the targeted index. Lists the specified schedule for when the indexing will occur. Lists the next scheduled run. This is N/A for ASAP schedule with no recurrence. Status includes: Pending Incomplete (Not fully defined) Active (Waiting for next run) Inactive (Disabled from UI) Finished Expired (Does not reach end date or complete "run once") Editing a collection activity You can edit collection activities in Data Protection Search. Procedure 1. Click the activity name. A summary of the activity with a list of all associated jobs is displayed. 2. To modify the collection activity, click the edit icon ( ). The wizard for the activity launches. 3. Make the required changes, and close the wizard. A completed activity cannot be modified. Example 1 Actions for lack of activity Some of the possible actions for lack of activity, include the following: remove run now enable disable Editing a collection activity 75

76 Collections 76 Data Protection Search 1.1.x Installation and Administration Guide

77 CHAPTER 8 Jobs This section includes the following topics: Jobs overview Jobs views...79 Data Protection Search job types Jobs 77

78 Jobs Jobs overview In the Jobs section of the Data Protection Search UI, you can view complete details and status for Data Protection Search collection, background, and system maintenance jobs. Index Admins create collection activities, which can be recurring. Every time a collection activity runs, a job is created. One collection activity can have many jobs. There are background system jobs, that monitor the system components, refresh client lists and backups, and run garbage collection to remove expired backups from the index. Also, there are jobs that are triggered from the Search UI, including Restore, Download, and Query. Figure 13 Jobs UI By default, collection, restore/download, and long query jobs are in the jobs list. System jobs are hidden to keep the list free from unnecessary information, and display only the most important job types. System jobs can easily be added to the list by clicking the Show system jobs icon on the toolbar, or by enabling System in the Jobs filter window. Index Admins are only able to view Index jobs (both metadata only and full-content). Available options for Jobs Actions available from the taskbar on the right side of thejobs UI, are listed in the following table. 78 Data Protection Search 1.1.x Installation and Administration Guide

79 Jobs Table 18 Available actions for Jobs Available action Stop ( ) Description Stop long running, or very large jobs such as full-content index that sends many requests. It stops any further requests from being sent to the Source. You can stop job requests that are in the queue. However, it is not possible to stop requests that are already being processed on the backup server. Filter ( ) Show system jobs ( ) Refresh ( ) You can filter on job type, status, the user who triggered the job, and activity name. Any job filters enabled here are active for the duration of the current session. If you return to the window and filters are enabled, the string Filtered results. Click the filter icon in the taskbar to remove or change the filter. is displayed above the list. Use Reset filter to return to a full view of all jobs. Refresh the list of jobs and their status. Jobs views Jobs status has an icon overlay when you hover over any job in the list. Table 19 Jobs views based on Admin permissions Job type Admin permissions visibility DPSearch Admin Index Admin Index (both metadata only and full content) No Yes Restore Yes No System maintenance (includes garbage collection, source cache, and node status) Yes No Query Yes No The Jobs view is filtered by using the following criteria: Job type Job status Triggered by (user) Activity name The Reset icon in the taskbar, resets the filter to the default (show all jobs). Tasks Each job is broken down into one or more subtasks, which are more granular work items that reflect one portion of the job. For example, when adding a backup server, a Jobs views 79

80 Jobs job is created to register agents for that backup server. Since every Data Protection Search Worker node must register its own agent, the job results in a separate task for each Data Protection Search Worker. Similarly, when a collection activity job runs, there is a separate task for each backup/save set on every indexed client. For full content indexing, there can be more than one task for large backups. It is useful to have access to the more granular viewpoint that tasks offer, particularly for failed jobs, or even just to understand how much of a long-running job has completed so far. Therefore every entry in the Job list has a View Tasks link to view the lists of tasks for that job. Each entry in the task list includes a status and details specific to the type of task. For example, a collection job shows details of the backup set, backup client, and backup server. Statistics are shown, indicating how many items have been scanned, processed, succeeded, failed, or are duplicates or updates. And the Worker that processed the task is identified, which can help in finding the correct log file to troubleshoot an issue. If a task failed, hovering over the icon shows an error message indicating the reason for the failure. A toolbar to the right of the list provides the option to refresh the list, or to return to the jobs list. It is also possible to filter and/or sort the list of jobs. Data Protection Search job types The Jobs section of the Admin UI provides information and status for indexing, search related, and system jobs. Indexing jobs When a collection activity is created, there are a number of scheduling options available. The collection can be scheduled to run immediately (ASAP) or at a future date, and can also be scheduled to recur, for example, daily, or weekly. Every time the collection activity runs, a job is created to process it. Clicking the activity name displays a list of all jobs that ran for the activity. In the Jobs View, statistics are listed for each indexing job, including: The start time The end time The number of items that are processed, succeeded, and failed Successful items are further broken down by the number of duplicate items, and updated items: Duplicate items are unchanged, and appear in multiple backups Updated items are initially indexed for metadata only, and later updated to be indexed for full content For activities with many clients, and clients with many backups or files, it is not unusual for indexing, and stats to take a significant amount of time. Indexing takes time to compile an internal list of files in the backups, and then divide into tasks. Also, it takes additional time to query backup servers for the information. Search related jobs Using the companion search web UI, Search Admins can complete actions on selected search hits. The actions include downloading and restoring search hits, and creating long running queries filtered by the backup date. These jobs can be tracked in the Search UI, and are also listed in the Jobs View in the Admin UI. System jobs To maintain the system, Data Protection Search runs background jobs at regular intervals. These system jobs include checking the status of Worker nodes, configured 80 Data Protection Search 1.1.x Installation and Administration Guide

81 Jobs backup servers, and garbage collection activities. Garbage collection removes files from the index that no longer exist on the backup server. Status jobs run hourly, garbage collection run jobs daily, and a special garbage collection reconciliation job runs monthly. The garbage collection reconciliation job synchronizes the backup information stored in DPSearch with the information in the backup servers. By default, system jobs do not appear in the Jobs View. To see them, check the System settings icon ( ). Data Protection Search job types 81

82 Jobs 82 Data Protection Search 1.1.x Installation and Administration Guide

83 CHAPTER 9 System This section includes the following topics: Data Protection Search System...84 System 83

84 System Data Protection Search System Worker Nodes The System section of the Data Protection Search Admin UI monitors, and provides information on all Data Protection Search workers and Elasticsearch nodes. When you open System, the Worker node, and CIS node entries are minimized. View the section by clicking the arrow below each one. The Worker node header bar is color coded according to status: Green (Good) Yellow (Warning) Red (Retired) The values that are displayed in the Worker Nodes section are gathered at hourly intervals, and therefore they might not be current. Specifically, memory and CPU utilization appear to be unusually high after a system restart. The following figure illustrates the Worker Nodes section of the System UI. 84 Data Protection Search 1.1.x Installation and Administration Guide

85 System Figure 14 Worker nodes System The Threshold (hard coded) and Current usage are listed for the following: CPU: >90% / >50% Memory: >90% / >60% Disk space: >90% / >60% The value in Current is color coded for when it is near or exceeds the acceptable threshold value. If enabled, notifications are sent when thresholds are exceeded. Notifications are also listed in, and can be viewed from System Notifications on the Data Protection Search Admin dashboard. Components The current versions of the following components are listed: Worker Node Admin web Admin API Search web Worker Nodes 85

86 System Search API The versions vary depending on patch levels and/or applied hotfixes. Agents The list of installed Agents and the status for each Avamar and NetWorker backup server is shown. CIS Nodes The following figure illustrates the CIS Nodes section of System. Figure 15 CIS nodes Elasticsearch/CIS collects the Threshold (hard coded), and Current usage information: CPU Usage: >90% / >50% OS Memory Used: >90% / >60% JVM Heap Used: >90% / >60% The value in Current is color coded for when it is near, or exceeding the acceptable threshold value. 86 Data Protection Search 1.1.x Installation and Administration Guide

87 System Notifications are sent when thresholds are exceeded. Notifications are also listed in, and can be viewed from System Notifications on the Data Protection Search Admin dashboard. CIS Nodes 87

88 System 88 Data Protection Search 1.1.x Installation and Administration Guide

89 CHAPTER 10 Options This section includes the following topics: Data Protection Search Options Configuring notifications Options 89

90 Options Data Protection Search Options The Options section of the Data Protection Search Admin UI provides the ability to configure and modify system and search options. The following figure illustrates the Options section of the Admin UI. Figure 16 Options The following table lists the available options. Table 20 System and search options Option Search Options Session Options LDAP Options Description Set the max hits to restore to limit the number of files that can be restored in one session. The range is from 101 to 5000 files. Set the session inactivity time out for both the Admin UI and the Search UI. The default is 1 hour before a login is required. You can set the range from 3 minutes to 24 hours. Modify the LDAP users/options (host) specified at deployment. 90 Data Protection Search 1.1.x Installation and Administration Guide

91 Options Table 20 System and search options (continued) Option Notification Options Index Options Description Enable notifications and set SMTP options. Set the value for the following: File types to exclude from FCI (list the file name extensions to skip during full content indexing) Number of replicas Apply Replicas settings to existing indexes The following table lists jobs and activities that trigger notifications. Table 21 Supported notifications Activity Restore operation Collection job (metadata or fullcontent) Index state Backup server state Notification trigger The notification is sent to the Search Admin, when a restore operation completes, succeeds, fails, stops, or times out. The notification is sent to the Admin that created the activity or forced a Run now, when a collection job completes, succeeds, fails, stops, or times out. The notification is sent to all Index Admins, when an index changes state (health, warning, or error). The notification is sent to all DPSearch Admins, when the backup server changes state (healthy or disconnected). Configuring notifications You can configure Data Protection Search to send notifications to specified SMTP users and hosts. The following figure illustrates the Notification Options to configure. Configuring notifications 91

92 Options Figure 17 Configure notifications Procedure 1. Click to toggle notifications from OFF to ON to configure notifications for Data Protection Search. notifications are disabled (OFF) by default. 2. Configure the following for notifications: SMTP (IP of the SMTP host) Port (Set an appropriate port, typically 25 or 587) SMTP User (An account authorized to connect to the SMTP service) Password Notifications are always enabled, and can be viewed in the Notifications section of the dashboard. These options relate only to whether notifications are sent. 3. Click Validate. Results Click to select the notification and view its details. Notifications take effect 15 minutes after this setting is enabled. 92 Data Protection Search 1.1.x Installation and Administration Guide

93 CHAPTER 11 Performing Searches This section includes the following topics. Search overview Search performance factors Search by using keywords...95 Search criteria management Using Lucene to search...99 Restore files in Data Protection Search Search results Search criteria management Search performance factors Performing Searches 93

94 Performing Searches Search overview Data Protection Search is a Google like search that performs full content index, or metadata only index searches. Data Protection Search uses keyword and Lucene queries. Search results include a summary of certain fields for the file, such as filename, location, backup client. If a keyword or phrase is found in the file name or path, it is highlighted. If a file is indexed for full content, and a keyword or phrase is found in the body of the file, a snippet of the full contents is returned. The snippet displays, and highlights the words around the matched text. The following figure illustrates the Search section of the Data Protection Search UI. Figure 18 Search UI Search performance factors By default, the search is sorted by Relevance (score), which is how well the hits match the search criteria. The speed at which search results are returned is dependent on a number of factors, including the following: How many items are in the indexes being searched If the Index filter is used to restrict searches to a specific index or indexes The number of configured Index Data nodes More nodes to distribute processing increases search performance The number of simultaneous queries Multiple Index Data nodes enhance performance by distributing the processing across the nodes 94 Data Protection Search 1.1.x Installation and Administration Guide

95 Performing Searches Search by using keywords Replicas can speed up performance If an index is static, performance improves, only in a multi-node cluster If collections are running for indexes during a search operation Collections can slow down some types of queries Search scope Search filters narrow the search scope, and improve search performance Also, a search that is applied to a billion items is slower than one with millions of items. Filters can be combined for greater benefit. For example, all.jpg files modified in the last year from client "my-sles-client" Use the following procedure in Data Protection Search to search by using keywords: Procedure 1. Type the keyword in the search dialog box, and click Search. Figure 19 Search dialog box Search by using keywords 95

96 Performing Searches Results are returned with the keywords highlighted in both the file name and the contextual snippet (only for full content index) of the original text located underneath the file name, location and item specific metadata. 2. If it is available, click Preview to view a text representation of the original file which is pulled from the index. Full content indexing provides all the text without the original application formatting. A near real representation of the file is displayed with enough data to identify the file. Image files have a thumbnail instead. Figure 20 Example preview It is not be possible to create a preview for all files. Sometimes, a preview is not available. If a full content indexed file contains less than 2 MB of text, the preview includes the original formatting, where possible. If there is more than 2 MB of text, only the text itself is in the preview, without formatting. Embedded images are never in the preview, regardless of file size. If the original file type is an image, the preview might be available with a smaller size of the picture. 3. To download the entire file (full access permissions are required), click the file name. This action retrieves the file from the backup server to the Data Protection Search node. The View Jobs dialog box can be used to monitor progress of the download. Once the download is complete, there is a link in the View Jobs dialog box for file to be downloaded to the client from which the web browser is connecting. Most browsers provide the ability to open or save the file. If you open the file, it is opened in a registered application for that file type, if one exists. 4. To refine the search and narrow the results, click Add Remove Filters. 5. Click Search, or click Save and Close in the filter form. The results are displayed with the strongest matches listed at the top. 6. You can click another item from the Sort By (DESC) filter to change the criteria on which the results are sorted. Relevance (sort score) is the default. 96 Data Protection Search 1.1.x Installation and Administration Guide

97 Performing Searches Figure 21 Search filters Search criteria management To narrow and reduce the time that is taken to return results, add search criteria. The following lists the available Search Criteria: Index Platform Server Client File Name File Type Size Last Modification Date Backup Date Location Unindexable Content Index Lists the indexes that the currently logged in Search Admin has access to. For example, searchadmin1 might have access to index 1 and 2, and searchadmin2 might have access to indexes 2, 3 and 4. To remove indexes from the search, uncheck them. Platform The following filters are available for the Platform criteria: All (default) Avamar NetWorker Server Specify the name of an Avamar or NetWorker server. There is a visual filter available for this option. Search criteria management 97

98 Performing Searches Client Select a backup server and then specify the name of an Avamar or NetWorker client. File Name The following table lists the available filters for the File Name criteria. Table 22 File Name criteria File Name value Contains (default) Begins with Ends with Description Specify a keyword contained in, beginning, or ending with the file name. Can include wildcard characters (*). File Type Narrow the search by File Type by typing a file name extension in the dialog box or using the pie chart icon. The pie chart breaks down the current search by frequency of file type. Only the top 10 most frequent file types in the search are shown. Each file type extension is listed below the pie chart. To narrow the search by file type, perform one of the following steps: Click the Pie chart icon. To eliminate that type of file from the current search, click a file type extension below the pie chart. To limit the search to only that type of file, click a file type extension in the pie chart. Figure 22 File Types Size Specify the file size (always in KB). To display a visual breakdown of the current search results, click the bar chart icon next to the size control. The frequency that 98 Data Protection Search 1.1.x Installation and Administration Guide

99 Performing Searches each range of file sizes occurs is represented in a bar chart. To rerun the current search filtered by that size range, click the bar. Last Modification Date Specify the date on which the file was last modified. Greater than, less than, or between also has visual filter. For the Last Modification Date criteria, jobs are listed in the View Jobs list as Query jobs. When the job completes, click Query Result to view the details. Backup Date To display a visual breakdown of the current search results for the backup date, click the bar chart icon next to the date control. Click a particular year to divide the results by month. To rerun the current search filtered by that date range, click a month. Using a wide range for the Backup Date criteria to search for a string can result in a long search window and negatively impact performance. Restrict the range to a single backup date for this filter to avoid performance issues. To enhance performance, jobs run in the background. The following table lists the available filters for the Backup Date criteria. Table 23 Backup Date criteria Backup Date value Greater Than Between Less Than Description Specify a backup date value for which the search returns hits between that date and now. To limit the search results to the specified time period, include two date values. Specify a backup date value for which the search returns hits from before that date. Using Lucene to search Location Specify the location of the file by using the physical path or just segments of the path. Unindexable Content Select Yes or No for this search criteria. DPSearch can find (or exclude) files that could not be full-content indexed. Before you begin A knowledge of Lucene is required to return accurate search results in Data Protection Search. This section explains how to perform advanced keyword searches by using Lucene in Data Protection Search. Using Lucene to search 99

100 Performing Searches Table 24 Example Lucene syntax Syntax examples field:value - field:"phrase " field :(value1 OR value2) _ missing_:field _ exists_: field field:qu?ck bro * field:quikc ~ brwn ~ foks ~ - "fuzzy" match field:"fox quick"~ 5 - proximity date:{" t00:00:00.000z" to " T00:00:00.000Z"} Description The field contains the value The field contains exact phrase The field contains one or both values The field has no value or is missing The field has any non-null value Use wildcards Words that are similar No more than 5 words between them Match any date in range date:{* to " T00:00:00.000Z"} Dates before 2012 date:" t00:00:00.000z" Specified date search age:> 10 All ages over 10 quick brown +fox -news The word fox must be present, news must not be Lucene does not support using an asterisk (*) symbol for the first character of a search string. Procedure 1. Type the Lucene query in the Search field. 2. To complete the search, click the search icon. Example 2 Example search strings einstein AND "marketing plan" AND platform:avamar AND size:[50000 TO ] AND type:pdf metadata_author:david Douglas", this format could be used for item specific metadata search, user could reference to the item specific metadata from any file in the search result by clicking More. name:king????er* -jpg Restore files in Data Protection Search Data Protection Search provides the ability to restore files back to their original location on the client from which they were originally backed up. Files can also be restored to an alternate location on that client, or to an entirely different client on the 100 Data Protection Search 1.1.x Installation and Administration Guide

101 Performing Searches same backup server. However, Only Search Admins with full access have permissions to restore files to a different client. Use the following procedure to restore individual, multiple, or all the files to their original location or an alternate location (within the restrictions of the backup server). Preview is available for search hits resulting from full-content indexed searches to verify that you have located the correct files to restore. Procedure 1. Click Restore below an individual file. 2. Click the checkbox beside multiple files, and then click Restore. 3. To select all files on the current page, click the checkbox above the search results, and then click Restore. When restoring multiple files, ensure all files come from the same backup server. Choosing to restore multiple files from different backup servers might cause the restore operation to fail. NDMP files can typically only be restored to a NAS device of the same type. For example, VNX to VNX, or Isilon to Isilon. Avamar NDMP files can also be restored to a file system client. Search results Different search results, controls, and status are displayed and available for full access Search Admins versus read-only Search Admins. Table 25 Search results for full access versus read only search admins Search result component Full access Search Admin Read only search Admin Application icon Displayed Displayed Pathname Active link to download/open the file No active link to download/open the file File name Displayed Displayed Last change date Displayed Displayed Client Displayed Displayed Plugin Displayed Displayed Contextual snippet Displayed if one or more keywords was found in the full content (body) of the file No contextual snippet displayed Platform Same for full, or read only access Displayed File size Displayed Displayed Backup server name Displayed Displayed The following is a list of the Data Protection Search controls for search results: Search results 101

102 Performing Searches Display for the number of matching results To process the query, displays the time that is taken to complete. To view restore jobs, displays a dialog box Change sort criteria Restore, Download long query jobs Display a Restore option for individual files Displays a Preview option for individual files (Preview is disabled for read-only Admins) Certain file types, including.log,.exe,.dll and.bin files are skipped for full content indexing, and are therefore not available to preview. Displays a More option to open a window to view additional details Checkboxes to select multiple items Top level checkbox to select all items Top level Restore option for multiple items Display the number of matching search results Restore, download, and query jobs for the logged in user and the respective results for downloads and queries are shown in View Jobs. The following table describes the specific information available when you click to view More for a file. Table 26 Details More Info item Description Index Information Index Name Full content indexed? Index Date The full name of the index True/False (Metadata index if false) The date the file was indexed Item Information Client operating system Created Date Title Author The operating system for the backup client The date the file was created and item specific metadata. Available only for Windows backups on the NetWorker server, excluding Linux and UNIX backups Only available for full content indexed and found in item-specific metadata Subject Backups All backups containing files are listed here. Listed for each file: Backup date Backup number 102 Data Protection Search 1.1.x Installation and Administration Guide

103 Performing Searches Table 26 Details (continued) More Info item Description Expiration date Full item-specific metadata information (list of name/values pairs) Search criteria management To narrow and reduce the time that is taken to return results, add search criteria. The following lists the available Search Criteria: Index Platform Server Client File Name File Type Size Last Modification Date Backup Date Location Unindexable Content Index Lists the indexes that the currently logged in Search Admin has access to. For example, searchadmin1 might have access to index 1 and 2, and searchadmin2 might have access to indexes 2, 3 and 4. To remove indexes from the search, uncheck them. Platform The following filters are available for the Platform criteria: All (default) Avamar NetWorker Server Specify the name of an Avamar or NetWorker server. There is a visual filter available for this option. Client Select a backup server and then specify the name of an Avamar or NetWorker client. File Name The following table lists the available filters for the File Name criteria. Search criteria management 103

104 Performing Searches Table 27 File Name criteria File Name value Contains (default) Begins with Ends with Description Specify a keyword contained in, beginning, or ending with the file name. Can include wildcard characters (*). File Type Narrow the search by File Type by typing a file name extension in the dialog box or using the pie chart icon. The pie chart breaks down the current search by frequency of file type. Only the top 10 most frequent file types in the search are shown. Each file type extension is listed below the pie chart. To narrow the search by file type, perform one of the following steps: Click the Pie chart icon. To eliminate that type of file from the current search, click a file type extension below the pie chart. To limit the search to only that type of file, click a file type extension in the pie chart. Figure 23 File Types Size Specify the file size (always in KB). To display a visual breakdown of the current search results, click the bar chart icon next to the size control. The frequency that each range of file sizes occurs is represented in a bar chart. To rerun the current search filtered by that size range, click the bar. Last Modification Date Specify the date on which the file was last modified. Greater than, less than, or between also has visual filter. 104 Data Protection Search 1.1.x Installation and Administration Guide