Oracle Endeca ITL and Expereince Manager High Availability Architecture

What are High Availability and Failover   High availability  as the ability for the system to continue functioning after the failure of... thumbnail 1 summary

What are High Availability and Failover

  •  High availability as the ability for the system to continue functioning after the failure of one or more of the servers.
  • Failover as one possible implementation of high availability. Failover is defined as the ability for client connections to migrate from one server to another in the event of server failure so client applications can continue to operate.

How to Setup ITL and XM Failover

Two Information Transformation Layer (ITL) servers which host Tools and Framework required achieving high availability and another for Disaster recovery (DR) if required. 
Only one ITL can be in active mode and remaining failover/backup should be passive mode . As per architecture, specific, in the diagram below, folders should be shared between primary and backup ITL servers using shared drive. DR generally hosted in the different data center so file system replication or other technique required for DR environment. The Shared drive can be setup during product and app installation.
Following points needs to be taken care of during installation
  • Installation path should be same in all environments to avoid any conflicts and issues.
  • Endeca Application Initialize services script required to run from all the ITL servers for the first time.
  • Make sure to copy same Dgraph cluster XML with same naming convention to avoid duplicate Dgraph instance
Find out block diagram below for HA understanding and setup

What needs to be shared between active and passive ITL servers:-

  1.  Tools and Framework state folder
  2. Endeca application Data folder
  3. CAS workspace shared folder

What happens if failover happens during mid-process 

  • Failover ITL would have to begin the process again if primary ITL Dgidx failed from Dgidx process.
  • Failover ITL would have to begin the entire process again if primary ITL forge failed.
  • Failover ITL would have to begin the entire process again if primary CAS indexing  failed.
Note :  ATG properties need to be changed to redirect indexing to failover ITL in the event of primary ITL fault, but best practice would probably be to restart the indexing process again.
For more Endeca related information visit:  http://www.ajayagrwal.com/

No comments

Post a Comment

Note: Only a member of this blog may post a comment.

Text Widget