Do we need to test anything for Local atom running on java 8 when Linux Operating System patch is applied?
Some times when I do minor adjustments in different Boomi processes, the customers/end-users ask me the same thing. And I do feel tempted to just say "No, it's just a small little thing. Don't worry. What could possibly go wrong with this?"
The truth is, and we all know it, everything can go wrong whenever something is changed.
I wouldn't spend hours testing after an OS patch update, but I would do a little test. Test a very simple test process, restart the atom and just use maybe 15 minutes to make sure it behaves normally.
So yes, I would do a basic test after a patch update. Just to be sure.
Thanks for the reply Leif.
I have never experienced any issue after an OS upgrade. That counts for several Linux flavours such as Amazon Linux, RHEL 6 and 7 and Ubuntu. Some best practices we use do minimise risks
A few days ago, I observed an 'atom crash' event. (first ever) Looking into it, I found that the system administrators did a shutdown of the windows server, before they take a snapshot of the server; then they bring it back up to do windows and vmware/etc updates; and then do another reboot.
Due to the 'atom crash', I had to do a further manual reboot in order for the atom to not think two processes were still in the middle of execution.
It appears that the administrators are doing an immediate shutdown, which to me is rather extreme since the atom may not have a chance to shutdown gracefully (especially if it is running a process). I think that the 'immediate shutdown' is what caused the crash and the two hung processes.
However, I don't know yet is the suggested 'time delay' should be applied when shutting down or reboot an 'Atom' server.
yea, this is about windows and not linux, but figured that it may still apply to routine maintenance work related to OS.
Retrieving data ...