Please keep the list posted. Am 29.10.2014 um 18:47 schrieb Disny Disny:
> Hello Reuti > this is the output of qhost and qstat -f but i don't know what it means so > i'm hoping you can help > > kind regards.. > > root@sgemstr:~# qhost > HOSTNAME ARCH NCPU NSOC NCOR NTHR NLOAD MEMTOT > MEMUSE SWAPTO SWAPUS > ---------------------------------------------------------------------------------------------- > global - - - - - - - > - - - > gcl1 lx-amd64 4 1 4 4 - 3.8G > - 6.7G - > gcl2 lx-amd64 4 1 4 4 - 3.7G > - 3.8G - > gcl3 lx-amd64 4 1 4 4 - 1.9G > - 6.7G - > shdwgcl4 lx-amd64 4 1 4 4 - 3.8G > - 3.8G - > root@sgemstr:~# qstat -f > queuename qtype resv/used/tot. np_load arch > states > --------------------------------------------------------------------------------- > all.q@gcl1 BIP 0/0/4 -NA- lx-amd64 au > --------------------------------------------------------------------------------- > all.q@gcl2 BIP 0/0/4 -NA- lx-amd64 au > --------------------------------------------------------------------------------- > all.q@gcl3 BIP 0/0/4 -NA- lx-amd64 au > --------------------------------------------------------------------------------- > all.q@shdwgcl4 BIP 0/0/4 -NA- lx-amd64 au This looks like there is no communication between the qmaster and the execds. Checking the output of: $ ps -e f shows the `sgemaster` resp. `sgexecd` running on the systems? Do you have a firewall in place? Maybe the port 6444 and 6445 needs to be opened. -- Reuti > > ############################################################################ > - PENDING JOBS - PENDING JOBS - PENDING JOBS - PENDING JOBS - PENDING JOBS > ############################################################################ > 4 0.00000 Sleeper root qw 10/23/2014 09:20:09 1 > root@sgemstr:~# > > > On Thursday, October 23, 2014 6:38 PM, Reuti <re...@staff.uni-marburg.de> > wrote: > > > Please check in `qhost` resp. `qstat -f` the state of the machines, i.e. > whether the execd can be reached by returning a suitable value for the > machines. - Reuti > > Am 23.10.2014 um 17:35 schrieb Disny Disny: > > > Yes during the exec installation it added a startup script but is there > > other startup i need to add to it manually?? > > > > > > From: Reuti <re...@staff.uni-marburg.de>; > > To: Disny Disny <disny.wo...@yahoo.com>; > > Cc: grid Engine Mailing List <users@gridengine.org>; > > Subject: Re: Queue instances dropped > > Sent: Thu, Oct 23, 2014 3:29:58 PM > > > > Am 23.10.2014 um 17:23 schrieb Disny Disny: > > > > > > > I have a problem with Sge ..after installing the cluster everything > > > wotked fine but when i shut down the pcs and in other time i start them > > > and try to submit ajob i got this message : > > > queue instance "all.q@gcl2" droped because It is temprerly not available > > > > > > queue instance "all.q@gcl3" droped because It is temprerly not available > > > > > > queue instance "all.q@shdwgcl4" droped because It is temprerly not > > > available > > > > > > queue instance "all.q@gcl1" droped because It is temprerly not available > > > all queues are dropped because of overload or full. > > > I appreaciate any help. > > > > > > Are the execd's running on the ndoes - maybe they need to be added to your > > startup mechanism to do it automatically in case you shutdown the machines? > > > > -- Reuti > > > _______________________________________________ users mailing list users@gridengine.org https://gridengine.org/mailman/listinfo/users
