Heya,
Can you gist the logs anyhow of the part where it terminates itself? Can
you check if you see an hs_err.pid file? if so, can you gist it? Can
you try and run ES without the mongodb river and see if it happens?
On Tue, Jun 5, 2012 at 6:10 AM, Kara karine.marabyan@aero.org wrote:
The service wrapper is going to restart the jvm correct? But at that
time I'll be dropping data that mongodb streams to ES.On Jun 4, 3:58 pm, Jörg Prante joergpra...@gmail.com wrote:
OK, so I think it is the segmentation fault of the JVM that is causing
the
trouble. Maybe running ES under the service wrapper helps, chances are it
will write some messages in the wrapper.log. There should be crash files
(hs_err_pid...log). If there are no JVM crash files, a good idea is to
enable core dumps (ulimit -c) and examine those dumps by gdb if a
segfault
happens again.
Another hint is to disable the sigar libs, but I really don't know if
sigar
works reliably on OpenStack virtual instances.Jörg
On Tuesday, June 5, 2012 12:40:51 AM UTC+2, Kara wrote:
Somebody else started a elasticsearch server on his machine without
the mongodb river. He terminated it to troubleshoot this problem
first. I don't see anything in the log file now but it still
terminates itself after awhile. I have to say that I am leaving the ES
server inactive for 1 hour and 15 minutes then it terminates. Next I
will try to keep the ES server busy to see if it terminates itself
even when it's busy.- Hide quoted text -
- Show quoted text -