..
>
> in reality these jobs # are not sequential.
>
> my problem/question is how do I cancel/qdel all of Team B's jobs easily
> It doesn't appear you can delete by job name
You can, even with wildcards:
$ qdel "*TeamB*"
The explanation of "wc_job_range_l
uld be a `mpihello.c` where you put an almost endless loop inside
and switch off all optimizations during compilations to check whether these
slave processes are distributed in the correct way.
Note that some applications will check the number of cores they are running on
and start by OpenMP (not
> Am 11.08.2016 um 12:14 schrieb Ulrich Hiller <hil...@mpia-hd.mpg.de>:
>
> Dear Reuti and Hugh,
>
> thank you for your quick replies.
>
> I have changed control_slaves to TRUE and job_is_first_task to FALSE,
> but it did not help.
>
> on my side:
>
el job may be delayed by one
to two minutes (until they face a timeout).
-- Reuti
> Am 10.08.2016 um 21:15 schrieb Ulrich Hiller <hil...@mpia-hd.mpg.de>:
>
> Hello,
>
> My problem: How can i make gridengine not to use ssh?
>
> Installed:
> openmpi-2.0.0 - config
he slaves (or even locally on the master node of the parallel
job) at times of MPICH(1) (will it allow [n] or [n-1] calls, also to check
proper programming by this). Nowadays Open MPI will start only one daemon per
exechost by `qrsh -inherit ...` and the granted limit won't be reached anyway.
Am 10.08.2016 um 21:35 schrieb Jaime Huerta Cepas:
> Thanks Reuti,
> the problem I have is that the queue system is used to handle jobs submitted
> through a web service. Each job is an array job, and if 10 web users submit a
> job, the last one in submitting has to wait un
'switch' where i can let qsub run _not_ over ssh?
>
> If is is of interest, the output of 'qconf -sp orte' is:
> pe_nameorte
> slots 999
> user_lists NONE
> xuser_listsNONE
> start_proc_argsNONE
> stop_proc_args NONE
...` on a per job basis (could even be put in
a JSV)
b) the setting "max_aj_instances" in SGE's configuration `qconf -mconf`
HTH -- Reuti
> thanks a lot in advance,
> -jaime
> ___
> users mailing list
> users@g
nly the
given hosts in the file you provide here)
- export OMP_NUM_THREADS=5
* The format is explained in `man sge_pe` start_proc_args "$pe_hostfile".
-- Reuti
PS: Maybe a prolog could also alter the provided $SGE_JOB_SPOOL_DIR/pe_hostfile
directly - I never tested it to alter it but
Then h_vmem will be reserved and at one point your jobs will start essentially.
-- Reuti
>
> Regards,
> Sudha
>
> The information contained in this electronic message and any attachments to
> this message are intended for the exclusive use of the addressee(s) and may
itching it off on a BIOS level will also change the internal operation of the
pipeline in the CPU as there are no threads to handle any more. IIRC there was
something about it on the Beowulf list.
-- Reuti
___
users mailing list
users@gridengine.org
https://g
t in the queue definition. This is
not related to a particular job setting. How would you change the priority for
a job?
-- Reuti
> Thanks in advance,
> Julien
> ___
> users mailing list
> users@gridengine.org
> https://gridengi
mountpoint? Maybe it was (re)mounted in read only mode - can you
create a file in /opt/gridengine/default/spool/qmaster by hand?
-- Reuti
___
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
to these tools.
-- Reuti
> Am 05.07.2016 um 13:18 schrieb Nacheva, Eva <eva.nach...@kcl.ac.uk>:
>
> Hello,
>
> I am using the qacct command for sge. However, as I don't need all the
> information that the command provides, I have been wondering if there is a
> way to tailor
Hi,
> Am 04.07.2016 um 15:14 schrieb Nacheva, Eva <eva.nach...@kcl.ac.uk>:
>
> Dear Reuti,
>
> In more detail, we are trying to use the prolog and epilog to gather
> information about each job and send this information to a database which is
> connected
ob during submission time?
-- Reuti
___
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
Hi,
> Am 01.07.2016 um 17:01 schrieb Jerome <jer...@ibt.unam.mx>:
>
> Dear Reuti
>
> Le 29/06/2016 à 16:34, Reuti a écrit :
>> Hi,
>
>> Am 30.06.2016 um 03:39 schrieb Jerome :
>>
>> Dear Reuti
>>
>> Thank's for your answer. I use
Hi,
> How ?
>
> To add the node, I think it is:
> qconf -ae $exechostname
>
> But how to add it to host list without any interactive (term/display) way (I
> am doing it remotely through SSH or salt).
$ qconf -aattr hostgroup hostlist node01 @intel2670
reuti@login modified
Hi,
> Am 30.06.2016 um 03:39 schrieb Jerome <jer...@ibt.unam.mx>:
>
> Dear Reuti
>
> Thank's for your answer. I use SGE for a while, and don't realize what you
> explain here with -now yes
It specifies whether the job must start immediately, or can start at any la
ial?
Interactive is more like immediate, and you can run an interactive job in a
plain batch queue with "-now no", resp. a batch job in an interactive queue if
you request "-now yes".
Note that parallel jobs are unrelated to this setting and even with qtype NONE
they will run as l
Hi,
Am 29.06.2016 um 11:28 schrieb Ueki Hikonuki:
> Hi Reuti,
>
> I tried your example, but since I'm not familiar with AR, I couldn't achieve
> what I wanted.
>
> Finally I came up with the following Ruby script, which waits until a
> specified job
> becom
still thinks they are
present there, you can use (as SGE manager):
$ qdel -f
to adjust the bookeeping.
-- Reuti
> Any advise would be greatly appreciated.
>
> Regards,
> Sean Smith
>
> ___
> us
the job requirements, it
would be better to implement it in another way: by using a forced boolean or
string complex with an attached urgency.
Special users can then request this complex (either by theirselfs or by an JSV
automatism) which must also be attached to a queue of your choice
t depends whether you choose classic spooling or berkley DB during
installation. Although changing the files by hand in the frmer case my work in
some cases, the best is to use the `qconf` command to change certain settings.
Otherwise changes may not be reflected instantly.
As these files are onl
can be requested though.
> Previous SGE versions used to have the option "qsub -l exclusive..", but
> 6.2u5p2 does not.
This is a different issue but nevertheless: did you define a complex
"exclusive" of with "relop" = "excl" and attached it to
ce depending
on the time the USR2 arrives, there is a forked sleep 1 or not which will use
the default behavior. You can try:
while (trap - usr2; exec sleep 1); do
-- Reuti
> Any idea ?
>
> Best regards,
> Julien
> __
and would like to restrict users from viewing others
> queue\nodes that are not assign to them(for ex: qstat\qhost commands)?
No. There is no such feature in SGE. You can use a wrapper, but users can
bypass it once they got the knowledge.
-- Reuti
__
message: “Unable to run job:
> failed receiving gdi request. Exiting”
Are all machines running the same version of SGE? Do the jobs later succeed on
exactly the same nodes they failed before?
-- Reuti
> But the job runs successfully when it is seen later with qstat.
>
> We tri
No, the prolog and epilog are run on their own like the jobscript. The
jobscript isn't a child of the prolog.
For such a behavior you would need a starter_method.
Still on vacation - Reuti
Von meinem iPhone gesendet
> Am 08.06.2016 um 12:00 schrieb Chris Dagdigian <d...@sonsor
qconf -sq all.q
...
pe_list smp
-- Reuti
> Or did i misunderstand your question?
>
> Kind regards, ulrich
>
>
>
> On 04/28/2016 07:38 PM, Reuti wrote:
>> Hi,
>>
>> Am 28.04.2016 um 19:18 schrieb Ulrich Hiller:
>>
>>> D
ccounting_summary FALSE
Was this PE also attached to any of your queues?
-- Reuti
>
> Then i created i simple file:
> ~> cat teste.ttt
> #!/bin/bash
> #
> #$ -cwd
> #$ -S /bin/bash
> #$ -o out.txt
> #$ -e err.txt
> #$ -pe smp 8
> /bin/date >
quot; should be mangled here, what is not done by default). I can't find it in
short-term, but I'll look it up.
Maybe it's done on some of your servers.
-- Reuti
> -Original Message-
> From: Feng Zhang [mailto:prod.f...@gmail.com]
> Sent: Monday, April 18, 2016 10:33 AM
>
ot /usr/sge/bin/lx24-em64t/sge_execd
root root \_ /bin/sh /usr/sge/cluster/tmpspace.sh
sgeadmin root \_ sge_shepherd-61837 -bg
...
-- Reuti
> Anywhere else I should be looking?
>
> Thanks for the time.
>
> -Original Message-
> From: Reuti [mailto:re...@s
.s...@gmail.com>:
> Hi,
>
> I tried to add four nodes into my grid, I have other 4 nodes working fine. I
> can see the nodes infotmation using qhost, but for a reason they don't submit
> the job.
The may be a static output, based on the information you added with `qconf -a
g the high priority queue they could have
their personal ".sge_request" in their home directory or even specific
subdirectories.
Another way would be to implement a JSV (job submission verifier) and attach
"-R y" only to certain jobs depending on some criteria.
-- Re
his bug is still here...
AFAICS this targets DRMAA_V_ENV. Using the DRMAA_NATIVE_SPECIFICATION my work
by setting "-v MY_VAR=foo=bar".
-- Reuti
> 2016-04-06 17:11 GMT+02:00 Julien Nicoulaud <julien.nicoul...@gmail.com>:
> Hi all,
>
> When submitting jobs via DRMAA with en
Am 16.04.2016 um 21:48 schrieb Smith, David [ETHUS]:
> I use sendmail. When you say the daemon, do you mean sendmail or grid engine
> daemon?
Mainly the `sendmail` one here, but essentially both should run as root.
-- Reuti
> Original Message
>
> Subject:
ed admin email address will get a copy where an error may show up
like "Can't open input/output file" or "No such file or directory" which might
point to an NFS bottleneck.
Can you pleaes check the emails of SGE sent to:
$ qconf -sconf
...
administrator_mail reuti
"MATCHING TYPES".
Despite that these values can be supplied just with a loop in `qconf -aattr
exechost`, some prefer to use a load sensor to get the correct values set
automatically.
-- Reuti
>
> Regards,
> Sudha
> The information contained in this electron
uld be changed/created to read:
$ qconf -sconf nodeXY
qlogin_command /bin/false
But there won't be a nice error message, it will just try to establish the
connection and fails essentially.
https://arc.liv.ac.uk/SGE/htmlman/htmlman5/remote_startup.html section "LOCAL
CONFIGURA
ot for all exec hosts. Does
> anyone know where this is controlled?
This depends on the used mailer. Postfix or sendmail?
Either a specific setting is made there to change the sender of the email
automatically, or maybe the daemon is running under a user account itself.
-- Reuti
>
>
n3380a02000-3380a0300"..., 1024) = 1024
> [pid 1048] write(2, "tils.so.1.3\n3380a02000-3380a0300"..., 1024) = 1024
> [pid 1048] read(9, " \n7fc6ba459000-7fc6bc00 ---p"..., 1024) = 1024
> [pid 1048] write(2, " \n7fc6ba459000-7fc6bc00 ---p"
chines, there are also several CPUs without HT capability.
-- Reuti
___
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
Hi,
> Am 13.03.2016 um 08:53 schrieb Yuri Burmachenko <yur...@mellanox.com>:
>
> Hello Reuti,
>
> We will try that, but we have also found another issue.
>
> We see that also our SoGE fails and failovers from master to shadow and
> vice-versa during the same t
> Am 10.03.2016 um 13:03 schrieb William Hay <w@ucl.ac.uk>:
>
> On Thu, Mar 10, 2016 at 11:15:58AM +0100, Reuti wrote:
>>
>>> Am 10.03.2016 um 10:45 schrieb William Hay <w@ucl.ac.uk>:
>>>
>>> On Thu, Mar 10, 2016 at 08:1
queue is needed. Would there be any flaw if
the internal check by GridEngine is done only after the JSV has run? This would
allow the JSV to correct any mis-submission and/or pseudo/dummy/alias to be
resolved, being it names of queues or complexes.
-- Reuti
signature.asc
Description: Messa
Hi Chris,
Am 08.03.2016 um 18:29 schrieb Christopher Black:
> Thanks for the reply Reuti!
>
> Sounds like some of the suggestions are moving limits out of RQS and into
> complexes and consumable resources.
Yep.
> How do we make that happen without
> requiring us
..
Here one could use a globale complex for each type of queue, as long as the
users specify the particular queue. One will lose the ability that potentially
a job may be scheduled to different types of queues, as long as resource
requests are met.
I can't predict whether this would improve anyt
> Am 08.03.2016 um 10:59 schrieb Yuri Burmachenko <yur...@mellanox.com>:
>
> Hello Reuti,
>
> See below:
>
> Job IDJob schedule time
> 97453 29-02-2016_03:18:55
> 97454 29-02-2016_03:18:57
> 563 29-02-2016_03:23:44
en references to this condition being fixed by deleting the
> job, but how do I do this? We use BDB spooling. This grid is running
> SGE 6.2U5.
Is the job still running? It looks like it finished already. Nevertheless: did
you try a `qdel -f `?
-- Reuti
___
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
Hi,
Am 06.03.2016 um 18:48 schrieb Yuri Burmachenko:
> Hello Reuti,
>
> Just last question, is adjusting MAX_DYN_EC is a disruptive action or not and
> is the effect immediate or restart required for SGE master?
All changes are live and won't interrupt any scheduling. In fact:
Am 06.03.2016 um 18:21 schrieb Yuri Burmachenko:
> Hi Reuti,
>
> No we didn't change that setting, but anyway I could not find it in `qconf
> -mconf`, see below the contents of this command:
Please have a look at `man sge_conf` about it. - Reuti
> execd_spool_dir
our systems including SGE master/shadow and its execution hosts.
To elaborate this. When it suddenly jumps to : what was the highest
JOB_ID which was recorded before that skip in the accounting file?
-- Reuti
> Perhaps something sets $SGE_ROOT/default/spool/qmaster/jobseqnum to “999
There is a setting of MAX_DYN_EC in SGE's configuration `qconf -mconf` , but
should default to 99. Did you increase it already beyond it to 950?
-- Reuti
> Does this mean that the limit is 950 grid nodes?
> Is it possible to increase this number?
>
> What are our options if we reach 950 grid no
ally in
`qconf -mc`) to each exechost and request this when submitting the reboot job?
Even one slot would be enough then to get exclusive access to each node.
-- Reuti
> script:
>
> ME=`hostname`
>
> EXECHOSTS=`qconf -sel`
>
> for TARGETHOST in $EXECHOSTS; do
>
>
es a being used
for further scheduling.
Are these quad-CPU machines without hyperthreading to gain a total of 56 cores?
-- Reuti
> - and I suppose that's exactly what the default "np_load_avg" does.. awesome!
>
> > we basically have 2 kinds of queue - a workhorse queue &q
never ever
> put our nodes in alarm mode, we use zabbix to monitor machine's health and we
> automatically take it out of the cluster (by disabling all of it's queues) in
> cases of "mess" (disk failures, out of space, mounting issues, stuff like
> that).
Are these in
tion - the alarm_threshold
would take care of this and avoid further jobs to be started on this machine).
If you have serial jobs only or parallel jobs which scale well, this isn't
necessary to be set.
(BTW: we use alarm_threshold only to put a machine in alarm state to avoid
further dispatc
quot; order
I never saw it before. And it's not an array task which got an illegal lower
boundary?
-- Reuti
> 02/16/2016 08:20:07|worker|nfs-2|W|Skipping remaining 352 orders
> 02/16/2016 08:20:07|worker|nfs-2|E|reinitialization of "scheduler"
>
> They persist over sc
| RT | start_time |
> FROM_UNIXTIME(end_time) | sbmt_time| (CPU/3600) |
> +-++-+-+-+---+---+---+
>
>
> | pangjx | 143320 |277.7775 | 145.3114 | 2015-07-23 21:56:51 |
> 2015-07-29 23:15:32 | 2015-07-23 21:56:40 |
277.777
ob), not to
mention parallel jobs where you can get a bunch of entries (depending on the PE
setting "accounting_summary").
What's the goal behind it? Do you want to start another job or start something
external?
Besides looking into DRMAA, there is also a small tool `qevent
e assigned and limited
on a exechost level to 4 (this works, unless the user foul the system and
request 0 for this complex, but a JSV could handle it). The complex could also
be assigned with an arbitrary high value on a cluster level and an RQS could
limit it on certain machines.
-- Reuti
> Sh
> Am 27.01.2016 um 04:40 schrieb Marlies Hankel <m.han...@uq.edu.au>:
>
> On 01/26/2016 11:22 PM, Reuti wrote:
>>> Am 25.01.2016 um 23:32 schrieb Marlies Hankel <m.han...@uq.edu.au>:
>>>
>>> Hi all,
>>>
>>> Thank you. So I
list while
> all other exechosts have NONE there. All of these job IDs belong to jobs that
> are either running fine or still queueing.
Ran this job before on node3 and were rescheduled already - are there any
`qacct` entries for them?
-- Reuti
> But I do not know from where this file
ee=sse4a, sse=sse4.1 and sse4.2 attached to the exechosts
complex_values line. Then one can request:
$ qsub -l "sse=sse4.?" ...
to get only the latter two. The complete documentation is in `man sge_types`
section "MATCHING TYPES".
-- Reuti
Am 25.01.2016 um 20:27 schrieb
The directory node17 will be
recreated automatically when the execd starts on node17.
-- Reuti
>
> Also, does anyone know what might have caused this error to appear in the
> first place?
>
> Best wishes and thank you in advance for your help
>
> Marlies
>
>
ight_urgency0.10
weight_priority 1.00
max_reservation 32
default_duration 8760:00:00
-- Reuti
> On Mon, Jan 25, 2016 at 11:25:53AM -0800, Christopher Heiny wrote:
>>
>> Hi all,
>>
>> We've been
n some nodes due to architecture/hardware constraints?
-- Reuti
> --
> Atte.
>
> Dimar González Soto
> Ingeniero Civil en Informática
> Universidad Austral de Chile
>
>
> ___
> users mailing list
> users@grideng
ition could be placed in each script, but this way it's
necessary to change it only in the checkpointing definition in case you want to
move it to a different location.
-- Reuti
*) This could be placed in a starter_method too.
___
users mailing list
users@g
> Am 17.12.2015 um 13:19 schrieb Steven Du <edgef...@gmail.com>:
>
> Hi Reuti/Joshua,
>
> I tried but failed. Here I wonder ( I am using 2011.11p1 version.):
>
> 1. If all GRID members have to be NFS shared spooling dir, such as
> $SGE_CELL/common. Otherwise, t
Maybe `mpirun` doesn't support/use Hydra. Although not required, the MPI
standard specifies `mpiexec` as a portable startup mechanism. Doesn't Intel MPI
also have an `mpiexec`, which would match the `mpirun` behavior (and doesn't
use Hydra)?
-- Reuti
> Am 17.12.2015 um 15:06 schrieb Gowt
call to `qrsh` automatically set up as you would need to start certain daemons
beforehand.
Does your version still need mpdboot?
Do you request a proper set up PE in your job submission?
-- Reuti
>
> Thank you for your time and help.
>
> Best regards,
> g
>
> --
> Gow
opt/software/$ARC/foobar
Having directories /opt/software/lx-amd64 and /opt/software/lx-x86 you will
always get the correct binary. I even use this to distinguish between amd64 and
em64t to get the correct binaries for each type of CPU although the jobscript
stays the same.
-- Reuti
> I wil
AFAIK it's fixed in SoGE. If you don't want to upgrade you can try an instant
fix:
$ cat trim.sh
#!/bin/sh
sed -i "/=()/s/.*/& ; }/" $SGE_JOB_SPOOL_DIR/environment
and:
$ qconf -sq all.q
...
prologsgeadmin@/usr/sge/cluster/trim.sh
-- Reuti
PS: One error is
uot;)
> fscanf(fp, sge_u32, _nr)
> guess_job_nr = guess_highest_job_number();
> job_nr = MAX(job_nr, guess_job_nr);
I wasn't aware that also the BDB is checked. For a fresh installation it should
be sufficient to fill the jobseqnum file with the proper value to start from.
-- Reuti
&g
adays, you could try to have
>> swap as large as the built-in memory.
>> The result should be, that any suspended job can be swapped out completely.
>>
>> -- Reuti
>
> So long I just lowered the default value of h_vmem (assigned when it isn't
> ordered explicit with
e.
Although I usually use only 2 GB of swap nowadays, you could try to have swap
as large as the built-in memory. The result should be, that any suspended job
can be swapped out completely.
-- Reuti
___
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
> Am 10.11.2015 um 17:33 schrieb Carl G. Riches <c...@u.washington.edu>:
>
> On Tue, 10 Nov 2015, Reuti wrote:
>
>>
>>> Am 09.11.2015 um 19:02 schrieb Carl G. Riches <c...@u.washington.edu>:
>>>
>>> On Sat, 7 Nov 2015, Reuti wrote:
>
> Am 09.11.2015 um 19:02 schrieb Carl G. Riches <c...@u.washington.edu>:
>
> On Sat, 7 Nov 2015, Reuti wrote:
>
>> Hi,
>>
>> Am 06.11.2015 um 22:03 schrieb Carl G. Riches:
>>
>>>
>>> We've configured a sharetree policy and would l
any custom script which tries to access
/tmp/9572476.1.preprod.q/pid - I can't remember that it's created by SGE in
this directory by default.
-- Reuti
> Can you please let me know the reason for these errors.
>
> Regards,
> Sudha
> The information contained in this elect
rue reporting=true \
> flush_time=00:00:15 joblog=true sharelog=21600
Were any jobs dispatched during the interval?
-- Reuti
> but nothing is being written to the reporting file. Is there something we
> missed?
>
> We are using Open Grid Scheduler/Grid
option too.
To ease the handling for the users, a JSV could add the exclusive resource in
case a particular PE is requested. It could also adjust the number of requested
slots from any value to the range "1-" to use the complete node, independent
from the number of built-in cores which
they're still text files. However you have to changes them through the
> qconf command. Fiddling
> aound with them directly isn't supported. they should be safe to read though.
As the page states: only for classic spooling w/o a DB. But a `grep` to get the
values is no big deal, as to c
export whatever this blank environment variable key is, it's
> core dumping.
I always tell my users to avoid using -V. Sometimes they made some changes in
the actual shell to the environment, submit a job and when the job starts after
3 days and crashes it's hard to investigate, as t
Ups - typo...
Am 31.10.2015 um 19:36 schrieb Leior Varon:
> Hi Yuri,
> I think it's best that you be the contact with the forum but I think it would
> be good to answer something similar to the below:
>
> Hi Reuti,
>
> Thank you for the reply.
> We also noticed tha
Am 31.10.2015 um 19:36 schrieb Leior Varon:
> Hi Yuri,
> I think it's best that you be the contact with the forum but I think it would
> be good to answer something similar to the below:
>
> Hi Reuti,
>
> Thank you for the reply.
> We also noticed that requesting &
E_ROOT/bin/lx-amd64/qsub -V -cwd -b y -q host_in_grid -pe testpe 1 xterm
>
> Any ideas why does this strange behavior occur? Is this some kind of a bug?
> How this can be resolved?
Unfortunately I have no idea, but I observed already in former versions that
instead of:
-l h=foo -q bar
it's
Am 30.10.2015 um 00:16 schrieb Lane, William:
> Is it possible to customize the kind of information included w/an SGE
> job-status email that is sent via the -m switch to qsub? I've noticed it
> includes (at least on our SoGE install):
>
> User = lanew
>
and
end time of a job based on the job number parsed from the subject?
Besides race conditions: one could parse the accounting file.
-- Reuti
-Bill L.
From: users-boun...@gridengine.org [users-boun...@gridengine.org] on
behalf of Daniel Gruber [dgru...@univa.com]
Sent: Tuesday, October 20,
Hi,
Are there local configurations overriding the global setting:
$ qconf -sconfl
-- Reuti
Am 20.10.2015 um 09:24 schrieb Lane, William:
> I've been trying to code a C program to process the SGE job-status email to
> determine if a job failed due to exceeding the h_rt limitation of a d
Is the $SGE_ROOT shared too?
The location of the spool directory for the exechosts can be checked in `qconf
-sconf` ("execd_spool_dir").
-- Reuti
>
> -- Shazly
>
> On Thu, Oct 15, 2015 at 11:45 AM, Reuti <re...@staff.uni-marburg.de> wrote:
> Hi,
&
output you may get.
-- Reuti
> Am 15.10.2015 um 14:52 schrieb Hatem Elshazly <hmelsha...@gmail.com>:
>
> Yes it is.
>
> Why do you think that the exec dirs weren't created? all the permissions and
> ownerships are granted.
> I'm using this script: inst_sge_sc to m
hared between them.
-- Reuti
> On Wed, Oct 14, 2015 at 12:20 PM, Reuti <re...@staff.uni-marburg.de> wrote:
> Hi,
>
> > Am 14.10.2015 um 05:42 schrieb Hatem Elshazly <hmelsha...@gmail.com>:
> >
> > Hello,
> >
> > I'm a debian user and I'm having
entials for connection
>
> Does this mean that this user should get SUDO privileges?
No. Do you start the daemons as root user?
-- Reuti
> On Thu, Oct 15, 2015 at 2:56 PM, Reuti <re...@staff.uni-marburg.de> wrote:
> The spool directory is created when the execd starts
They must run as root to allow switching to any user to run a job.
reuti@node:~> ps -eo user,ruser,command | grep sge
sgeadmin root /usr/sge/bin/lx24-amd64/sge_execd
> Am 15.10.2015 um 15:43 schrieb Hatem Elshazly <hmelsha...@gmail.com>:
>
> No. the daemon process is
$ qconf -sconf | grep gid_range
Are real groups occupying the same specified range and processes outside of SGE
use these too?
-- Reuti
>
>
> We run OGS/GE 2011.11p1 under Rocks 6.1.1. We had the same problem
> (sge_execd) with Rocks 5.x.
>
> Any pointers/hint/suggestions/etc w
ring to the Debian packages from their distribution? Then it's
better to ask on the Debian list, where the maintainer of these packages will
read it.
This list targets the configuration of SGE and compilation from scratch, but
not individual Linux distributions.
-- Reuti
>
> Any
ossible of course,
Do you submit jobs with "-notify" and/or a set s_rt?
--Reuti
> CPU = 00:00:00
> Max vmem = 10.229M
> failed assumedly after job because:
> job 187.1 died through signal KILL (9)
>
> I had thought an exit status of 0 indicat
Hi,
Am 24.09.2015 um 23:52 schrieb Lane, William:
> Reuti,
>
> 1.
> The exechost isn't the head node is it? We've always referred to our SGE
> clusters as having
> three types of nodes: submit nodes, compute nodes and head nodes.
It's better to think of machines as of
Hi,
Am 23.09.2015 um 21:18 schrieb Lane, William:
> Reuti,
>
> 1.
> If more than one compute node takes part in the compute ring, how does one
> determine
> which one is the exechost?
What do you mean by compute ring - a parallel job?
The exechost is the one where the job
301 - 400 of 1847 matches
Mail list logo