have very strange problem with my Dynatrace 6.5 installation and I don’t have any clue
why this problem occurs or what could be the reason for it.
6 hours (12AM, 6AM, 12PM, 6PM) a scheduled task (monitor) is failing (see
The failing task/monitor is not always the same but more or less a
random one; also it is not always the same plugin which is used. It is always just one job which fails –
I don’t see any pattern there. I also didn’t find anything in the log files
that could be helpful.
Does anyone have an
idea why this could happen? I know that it’s basically just guessing why this
could happen, but maybe, someone had something similar before. 🙂
Marius, I would also look in the task logs. These are located on the collector where the task is to be executed. If the log seems missing only for these specific executions, then that's indicative of the fact that the task could not run, leading credibility to Kyle's theory that it's a collector health issue.
I would also look in the Server.log, collector.log for information. An increased log levels may be necessary, if you're not getting anything out of the default levels.
i would also examine your schedules. Do you have a schedule that's defined to run things every 6 hrs?
thanks for the answers.
I solved the problem, now.
The problem was a Plugin which threw an exception if a certain connection could not be established instead of using the recommended "Status" object. Now, we used a script to autoconfirm already solved incidents using the rest api. As soon as the "self monitoring" profile contains no incidents anymore (because they are confirmed) the next plugin execution throwing an exception will cause a new execution failed incident. This effect does not happen as long there is an unconfirmed incident existing.