Andrey Kiryanov wrote:
> Hello,
>
> Maarten Litmaath wrote:
>>> of these runs the CE was under high load and we had lots of zombie
>>> processes
>>> that we had to clear by restarting the globus-gma daemon. Below is
>>> my job
>>
>> A new version of globus-gma fixing the zombie problem should appear soon.
>
> As I see that the problem has a bit larger scale than expected I have
> prepared a list of LCG-CE's that seem to be affected by bug #48588:
[...]
> nanlcg01.in2p3.fr:2119
[...]
> This list only includes CEs that have globus-gma running and allow dteam
> VO to run jobs. Site administrators of affected CEs are kindly requested
> to restart globus-gatekeeper service in order to get rid of the bug's
> effect (this has to be done only once).
Hello,
This morning I had this problem on nanlcg01.in2p3.fr and had to restart
globus-gma. As I recently updated lcg-CE on this machine, I would have
expected that this problem was corrected but apparently not.
Versions :
lcg-CE-3.1.28-0
glite-initscript-globus-gridftp-1.0.2-1
vdt_globus_jobmanager_lsf-VDT1.6.1x86_rhas_4_LCG-1
glite-initscript-globus-gatekeeper-1.0.0-1
vdt_globus_rm_server-VDT1.6.1x86_rhas_4_LCG-1
globus-job-manager-marshal-1.7.3-lcg
glite-security-voms-api-noglobus-1.8.8-2.slc4
vdt_globus_essentials-VDT1.6.1x86_rhas_4-9
vdt_globus_rm_essentials-VDT1.6.1x86_rhas_4-7
vdt_globus_jobmanager_condor-VDT1.6.1x86_rhas_4_LCG-1
vdt_globus_jobmanager_pbs-VDT1.6.1x86_rhas_4_LCG-1
globus-job-manager-marshal-client-1.7.1-lcg
vdt_globus_data_server-VDT1.6.1x86_rhas_4-7
globus-gass-cache-marshal-1.4.4-lcg
vdt_globus_jobmanager_common-VDT1.6.1x86_rhas_4_LCG-3
globus-gma-1.0.9-lcg
JM
--
------------------------------------------------------------------------
Jean-michel BARBET | Tel: +33 (0)2 51 85 84 86
Laboratoire SUBATECH Nantes France | Fax: +33 (0)2 51 85 84 79
CNRS-IN2P3/Ecole des Mines/Universite | E-Mail: [log in to unmask]
------------------------------------------------------------------------
|