r10 - 2017-03-20 - 15:36:07 - StevenBeardYou are here: TWiki >  Deployment Web > DeploymentPlanningAndDesign > ApproachesToImplementingHAAndDR > JazzIndexesBackupAndRecoveryConsiderations

Jazz indexes backup and recovery considerations

Authors: TimFeeney
Build basis: v4.0.x and 5.0.x Jazz applications in CLM and SSE

The Jazz CLM applications utilize a set of indices to implement its query and search services. Care must be taken to ensure these are backed up in a consistent state and restored propertly to recover from various errors, issues and failures that may occur

It is important to understand the structure and technology behind these indices, the correct procedure for backup and restore and any considerations for relevant failure scenarios.

Jazz indices at a glance

In the Jazz CLM applications, different implementations of the query and search services exist based on how their implementation architecture and what frameworks are used. Since CLM v4.0, all the applications use the Jazz Application Framework (JAF), however, how the application is constructed and the SDKs in use lead to different implementations of the query and search capabilities. The three primary implementations can be categorized as Item Query Service, JFS Query and Search Services and Fulltext Search Service. All applications make use of the JFS Query and Search Services implementation. CCM and QM also make use of the Item Query Service and Fulltext Search Service. This is important to know because the implementation used impacts how backup and restore is performed. In particular, it is the indexes based on a Fulltext Search Service implementation that need to be handled more carefully. For more complete details on the query and search service implementation see Query, Search and indexing technologies in CLM and Indices storage and management: Backup, recovery and recreation.

Backup and Restore

As of 4.0.5, JFS indices can be backed up online, while their application is running and as of 5.0.1, the Fulltext indices used by CCM and QM can be backed up online as well. For that reason, prior to 5.0.1, it is advised that all indices are backed up, along with their corresponding databases, while the servers are shutdown. Similarly, restore the indices and databases while the servers are shutdown.

See Backing up CLM for the recommended strategy for a comprehensive offline backup and restore of a CLM environment. In particular see the sections on backup of the JFS and Fulltext indices and temporal considerations when doing so.

Failure scenarios/considerations

Prior to 5.0.1, with the exception of the work item Fulltext indices, if Jazz indices are suspected of being corrupted, a backup could be restored and the applications will eventually 'catch up' the indexes to current state with no reindex operation required. Reindexes are costly and time consuming. A more desirable strategy is to perform nightly backups and if a failure occurs, restore from backup and let the indexes be caught up incrementally by the applications. Starting with 5.0.1, this can now be done with work item Fulltext indexes.

When a Jazz application fails or the indices are suspected of being corrupt, it would be good to run some sort of verification on the indices before performing any recovery process. This is now possible starting in 5.0.1. See Verifying indices.

Related topics: Approaches to implementing high availability and disaster recovery for Rational Jazz environments

External links:

  • None

Additional contributors: None

Edit | Attach | Printable | Raw View | Backlinks: Web, All Webs | History: r10 < r9 < r8 < r7 < r6 | More topic actions
Deployment.JazzIndexesBackupAndRecoveryConsiderations moved from Deployment.CLMIndexesBackupAndRecoveryConsiderations on 2014-03-10 - 09:16 by StevenBeard -
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Contributions are governed by our Terms of Use. Please read the following disclaimer.
Ideas, requests, problems regarding the Deployment wiki? Create a new task in the RTC Deployment wiki project