Hi Fabio,
do you have global config, qmaster_params, SHARETREE_RESERVED_USAGE
configured?
In this case sge_execd doesn't report the actual values, but
- wallclock * number of slots as cpu and
- requested memory * number of slos * wallclock as mem.
If a job has no memory request I would expect it to report 0 as mem.
Joachim
Fabio Martinelli wrote:
Hi Dave
I don't see errors for proc:
# strace /gridware/sge/utilbin/lx24-amd64/loadcheck 2>&1 | grep
--color proc
readlink("/proc/self/exe",
"/gridware/sge/utilbin/lx24-amd64/loadcheck"..., 4096) = 42
rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
open("/proc/cpuinfo", O_RDONLY) = 3
open("/proc/stat", O_RDONLY) = 3
open("/proc/loadavg", O_RDONLY) = 3
open("/proc/meminfo", O_RDONLY) = 3
open("/proc/stat", O_RDONLY) = 3
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
open("/proc/stat", O_RDONLY) = 3
num_proc 1
and please why if I do a qsub hard request on memory then I see qstat
properly reporting the live memory usage ?
hard resource_list: h_vmem=1G
...
usage 1: cpu=00:00:35, mem=35.00000 GBs,
io=0.01888, vmem=1.000G, maxvmem=1.000G
thanks again,
Fabio
On 7/19/11 12:38 AM, Dave Love wrote:
In case it's not clear, it should just be happening on any supported
platform. In case there aren't any better ideas, running
utilbin/.../loadcheck under strace might give a clue what's going wrong
(if it's somehow failing to read /proc, for instance).
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
