Solved, thank you.
It turned out I was able to start and keep the daemon up when I logged
in as root, but not if I did sudo -E (to export the environment);
Once I re-sourc-ed ${SGE_ROOT}/default/common/settings.sh, everything
got back to normal.
I bookmarked the website you sent, a great resource, thank you Rayson!
Best
Miro
On 05/19/2012 02:13 PM, Rayson Ho wrote:
The error message is the side effect, likely not the cause of qmaster
not starting - ie. when you don't enable the JVM thread, then you will
get that in the log when you shutdown the qmaster.
See the "qmaster or other Grid Engine daemons keep crashing" section
in the Trouble Shooting (written by me a while ago - it applies to
every version) and see if you can find out what's going on:
http://gridscheduler.sourceforge.net/howto/troubleshooting.html
Rayson
On Sat, May 19, 2012 at 4:55 PM, Miro Drahos<[email protected]> wrote:
Hi,
I have been happily running the gridengine, with qmaster installed on Fedora
16, until today, when I ran yum update and (among other things) gridengine
packages got updated.
Now I cannot start the qmaster daemon, I get
|E|jvm thread is not running
error in the qmaster's spool log
During installation some time ago I had chosen not to use jvm.
I tried reinstalling the qmaster, but to no avail -- the qmaster daemon
cannot start, it exits quietly, the jvm error in the log is the only
trace...
[One other sysadmin thing I have been doing today was installing LDAP, which
is why I ran yum update in the first place].
Thanks for any pointers
Best
Miro
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
