Pega Predictive Diagnostic Cloud™ (PDC) generates the OPS0014 event when PDC detects that updates to the search index are unacceptably delayed.
Reason for the event
Pega Platform™ periodically sends the state of the full text indexer queue to PDC. This information includes the date and time of the oldest pending update. PDC evaluates this data and generates an OPS0014 event if the oldest item has been waiting for more than two hours. PDC then creates a unique StaleSearchSuspect case for each application server on which the problem occurs. All subsequent OPS0014 events on the system are correlated to the same case.
The OPS0014 event also triggers a Search update latency notification to all recipients that subscribe to the event. For more information about notifications, see Subscribing to notifications in Pega Predictive Diagnostic Cloud.
Example message text
The Search Indexing update queue has entries that have been waiting for processing for several hours. It appears that there is unacceptable latency in index updates.
Recommended next steps
To resolve Search update latency issues, perform the following action:
Engage a subject matter expert to evaluate the state of the queue
One or more items in the FTSIndexer queue have been in the ready state for several hours.
Insufficient capacity might cause this issue, which you can address by enabling the FTSIndexer agent to run on more node types, and by reducing the latency between agent queue runs.
Alternatively, insufficient query performance might cause this issue. Because the full text index queue is subject to continuous inserts and deletes, the underlying database table might require compaction or cleanup. Check for performance alerts for the sppr_sys_reservequeitem_b stored procedure. Engage a database subject-matter expert to optimize the database table by using a command that is appropriate to the database management system, for example,
vacuum analyze pegadata.pr_sys_ftsindexer_queue.