· May 22, 2018

Permissions Change on 2017.2 vs 2017.1 When Shutting Down?

We upgraded from 2017.1 to 2017.2 today and noticed that when shutting down we must do it as an administrator.  Attempting to shut down as a non-admin result in a failure to shut down message.  We are curious if there was an issue with our upgrade or if there is a change that requires an administrator to shut down Cache in 2017.2.


Operating System is Windows Server 2012 R2.  We run Cache in a cluster using Windows Failover Cluster where we swing the drive from one machine to another when failing.  

Discussion (5)0
Log in or sign up to continue

Yes, we have a service account, that previously was not an administrator that was used to start/stop cache.  We installed using an administrator account in both instances.  The change seems to have happened when we upgraded to 2017.2 from 2017.1.  On 2017.1 we were able to start/stop cache without being an administrator.  Our service account was able to failover properly without being an administrator.  After the upgrade, the failover process failed when attempting to stop cache.  The CConsole.log does not show any errors, or anything related to shutdown which surprised us.  

Ultimately we have the issue solved by adding our service account as an administrator on the boxes, but we preferred to not do that as it seems to be overkill.  

We have opened a ticket in regards to this issue and will work it from that angle now.  

I appreciate your help, if we find a resolution I will come back and post it here in case others experience the same issue.

I don't see anything recently suggesting a similar problem or any changes in the product (though I may have missed something). I suggest filing a WRC issue for this for a more detailed investigation, but I'll give you my best guess here:

I suspect you typically start/stop the instance as a user that is not an Administrator. When you did the upgrade, you did so as an Administrator, and that left the instance running, started by an Administrator. The difference in which user started the instance likely caused the problem shutting down as another user. Given that, I'd expect that subsequent startups as a non-Admin user would allow for shutdowns also as a non-Admin user.

If this isn't the case, please provide the cconsole.log and the Caché SYSLOG ('do ^SYSLOG' from the %SYS namespace) to the WRC, and we'll be happy to help.