Dashboard “Top Shares by Usage” has started reading 0’s across board for a little while. I don’t know if it was something I did or an update. There is no error pop-up when refreshing dashboard and I do not know where logs are for this sort of thing. I recently updated mariadb for another service I have running on my server and was wondering if maybe Rockstor had some DB somewhere I blew away? everything else on dashboard appears to be working correctly. Ideas on where to start to look?
Thanks.
@Johnathan_Herndon Welcome to the Rockstor community.
If you are running stable channel updates and have disable quotas on the pool these shares belong to then that may explain it. The last released testing channel update (same as 3.9.2-0 stable, breaks all over the place when quotas are disabled. It is unfortunately a side effect of quotas disabled state. We use quotas to do these size calculations and if they are disabled they end up reading as zero.
You can re-enable quotas and after a while these figures should return, but of course there is then the performance hit of having quotas enabled. This is likely to be reduced significantly in kernel 5.2 onward so may become less of an issue.
Rockstor uses Postgresql so I think you are safe on that one.
This is not considered an error condition, just an inevitability currently. Ideally we shouldn’t have to disable quotas at all but currently it is not infrequently required for practical performance reasons, especially on large pools or if there are many snapshots.
But we do have a mouse over / tool-tip on this setting in the Web-UI, see the pics in the following post:
The main Rockstor log is:
/opt/rockstor/var/log/rockstor.log
Hope that helps and let us know if quotas disabled was the cause.
1 Like