How to monitor the rmngr process?

Question

1C

Hey!
there is 1s server 8.3.2.
installed on Ubuntu Server 12.04 LTS.
it is noticed that clients sometimes disconnect from it with an error: Server is unavailable (Does not respond, terminates abnormally or the port is occupied by another application).
telnet to the port works on the server itself.
free memory is full, the processor is idle, LA <1.
Questions:
1 How to determine the cause?
2 how to correctly monitor the freezing process and restart it?

Answer 1

I would like to learn how to monitor at least somehow .... not that competently =)
People out and under Win with similar problems toil: www.gilev.ru/forum/viewtopic.php?f=9&t=943
And what do you mean by "restart his"? The entire cluster manager?
In 1C, almost many problems turn out to be developer flaws and are fixed through communication with technical support, long promises of improvement, and the actual improvement. And in another way, only some kind of cardinal reconfiguration of the OS / modernization (which, however, can solve one problem and create three others), or personal experimentation, attempts to understand the causes of hangs or inventing your own workaround, such as creating an external monitoring process, which will be what - then restart there if the response from the server part 1s disappears.