We are intensive users of JIRA & Confluence and soon Bamboo, we have around 250 users.
I have noticed that performance and frequency of errors (SSL, Index, class crashes) are slowly but steadily increasing. And if there is an issue we often end up re-starting anyway.
Does it make any sense to have a regular re-boot/re-start of servers?
The last time we stayed up around 30 days - with issues mounting up as described above.
What kind of maintenance/re-starting schedule-strategy (in comparable user scenario) are you applying?
Better than trying to find a scheduled restart mechanism, try to avoid not needed restarts. I know that from time to time is good to free up some resources but JIRA should be stable enough to avoid restarts. Number of users do not only count when speaking about performance. There are other aspects that may be important. Try starting from looking at your infrastructure and JVM settings.
I do understand your view-point. But my question is rather, has anybody such a schedule in place?
I can just by cross checking monitoring see that there is a relation between performance and last re-start.
Its not that we have a lot of issues - rather the opposite.
Just want to see how the "general" community is seeing this. Of course maintenance and care on underlying infra and JVM is certainly a good hint. Consider that we have resources and expertise to have any eye on that.
In simple words - would you recommend a regular re-start - every week, month etc.... as a practice of maintenance?
I would not recommend a restart at all. You should really be investigating why there are problems and fixing the root cause. I would recommend a regular maintenance cycle - i.e. upgrading, and restarting during that. A number of my clients do have regular restarts in place, but in most cases, they have discovered that all they're doing is hiding a deep rooted problem. It's a bit like plastering over the cracks in the wall at home, only to realise too late that you've got a subsidence problem. Frequency of restarts depends on how often you're having the issue. As a temporary measure, establish when faults are occurring and simply schedule something a little more frequent. Then you can investigate your real issue.
Connect with like-minded Atlassian users at free events near you!Find a group
Connect with like-minded Atlassian users at free events near you!
Unfortunately there are no AUG chapters near you at the moment.Start an AUG