Hi, when logging calls I've been asked a few times by Symantec support staff whether I regularly restart or reboot my EV server.
Is this something that is unofficially recommended and if so what should the frequency be? We only have one EV server.
Cheers.
Karmakoma,
We restart all EV services nightly, on advice from Symantec. This resolved a number of issues we were getting, particulalry with errors appearing in the Event Logs.
We don't have any scheduled server reboots, although it probably gets bounced for other reasons about once a month...
Mick
Thanks Mick. Prior to or after backups?
We restart the EV services each night as part of our backup script. We don't restart MSMQ or the servers themselves.
Bern
..
The provided URL doesn't exist...
could you please confirm whether its recommended to perform the schedule reboot of EV 12.3 version as well?
Hi,
It is best practice to reboot occasionally. I reboot once a week all EV servers.
In future, please post new question, instead of adding to existing one. Creating a new one gets more attention. Also keep in mind links to ...symantec.com kb articles does not work anymore.
This is the link to the KB and documents site: https://www.veritas.com/support/en_US
I cycle the services nightly as part of my backup process - put them in read only mode, backup and then restart in R/W mode. I did read somewhere on the Symantec support site that you should stop EV services "before" rebooting the server because of an index corruption risk. Sure enough - that happened to me. Luckily it was not too far after the initial install, so the indexes only took a few hours to rebuild.
This happened to me also but i found that the reboot without stopping the services does not always cause the indexes to be damaged. In most cases the indexes dont actually get corrupt what usually happens is that the checksum.dat file in the index directory goes out of sync.
The fix for that is to stop the services again and then rename the checksum.dat to something else like Checksum.old.
When you restart the services a new checksum file is created.
If this works then the indexes are ok
If this fails to correct the issue then and only then do you need to either rebuild or you can restore from tape (or other backup media) and then do an update
This should bring the index online
