My suggestion would be to actually identify the root cause behind the obis crashes.Have you looked at what may be causing these crashes?
+1 to Joel. What's the root cause?
What do the log files say?
Are there core dumps?
etc. etc. etc.
That is the plan, but in the meantime I need to get the environment available and semi-stable at least.
The cause of the crashes seems to be this:
------------- log ----------------
[2019-08-05T23:55:19.529+02:00] [OBIS] [NOTIFICATION:1]   [ecid: 005ZqMkfYBk1vXaj9L6iKD0004U4004j2f,0] [sik: ssi] [tid: eb37700]  The cache polling delete statement failed for table .....
[2019-08-05T23:55:19.529+02:00] [OBIS] [NOTIFICATION:1]   [ecid: 005ZqMkfYBk1vXaj9L6iKD0004U4004j2f,0] [sik: ssi] [tid: eb37700]  The cache polling delete statement failed for table
In order for us to be able to help you, please post that in your initial question next time. We can't see your system or read your log files
But going back to my initial question - is there a way to increase number of automatic restarts for OBIS?
That's pretty much attacking the problem from the wrong end. Beating a dead horse 5 times instead of 2 won't make it magically well again.
I have systems which are running without a single restart. It isn't supposed to fail at all
What is going wrong and why. That's important. If you want you can always schedule a graceful shutdown and startup nightly with a cronjob and the stop/start commands.
Did you give up on this topic or are you still working on it?
For the benefit of the community it would help if - once concluded - you marked the corresponding answers as correct and helpful.