Implications of restarting EV services and DA Accelerator services
Hi
Environment is EV 9.0.2
We currently restart the following EV services nightly:
Enterprise Vault Task Controller Service
Enterprise Vault Storage Service
Enterprise Vault Shopping Service
We do this for historical reasons as we did it in our older EV8 environment, I believe it was to clear up MAPI sessions andkeep the system running more smoothly. Is this still necessary?
Does restart these EV services impact DA searches and exports? What's the impact?
Does restarting the DA Accelerator service impact DA searches and exports? What's the impact?
thanks
Hi Goatboy,
1. Restarting the EV Services : At the time of DA Export or DA Search usages EV Indexing and EV Storage Services (DLLs). If you restart the EV Services at the time of DA Search / DA Export the search / export will be affected. It means it will try to collect the data from ev server. If EV service restart will take more time than DA search / DA Export timeout or retrycount then those search / export will fail. Otherwise it will work smoothly.
2. Restarting the DA Service : Restarting the DA service (the Enterprise Vault Accelerator Manager Service - a.k.a., EVAMS) pause all running searches and exports to stop until the service is running again. Whenever service is started all pending search and export will start from the point whether they stoped.
Ajay is correct in that restarting the EV Storage Service during an export will cause items attempted to be exported to automatically retry. If we retry then enough times (5 by default) and fail, we fail those items and the export will have to be retried to retry those items again.
We recommend stopping EVAMS any time the EV Directory,Indexing or Storage service processes are stopped in order to effectively 'pause' any searches, reviews or exports while those services are down. We also recommend stopping EVAMS whenever the SQL Server hosting the DA databases is restarted as the communications channel between the DA and SQL servers would be dropped and not retried after the SQL Server is back online. This loss of the communications channel causes Event ID 158 errors to be thrown into the DA server's Symantec Enterprise Vault Event Log every 10 minutes for every DA Customer (including Custodian Manager).