[gridengine users] users Digest, Vol 51, Issue 10

Yuri Burmachenko yuribu at mellanox.com
Wed Mar 25 12:28:18 UTC 2015


thank you for your reply,

The issue is that it did work that way with SGE 6.1u6 - even when the reporting variable was not reported multiples of the load_report_time it still remembered its previous value and stored it.
Looks like in SGE 8.1.8 it drops reporting variables which are not reported during 'load_report_time' almost immediately.

Is there a way in SGE 8.1.8 to rise the timeout (ttl) for reporting variables which are not being reported for multiples of the load_report_time?

Regarding the suggested approach - it will still pass the value from the execution host to the Grid Master.
Is there a way not passing the value that frequent and instead reuse previous value for particular reporting variable?

Thank you!

-----Original Message-----
From: users-bounces at gridengine.org [mailto:users-bounces at gridengine.org] On Behalf Of users-request at gridengine.org
Sent: Wednesday, March 25, 2015 2:00 PM
To: users at gridengine.org
Subject: users Digest, Vol 51, Issue 10

Send users mailing list submissions to
	users at gridengine.org

To subscribe or unsubscribe via the World Wide Web, visit
or, via email, send a message with subject or body 'help' to
	users-request at gridengine.org

You can reach the person managing the list at
	users-owner at gridengine.org

When replying, please edit your Subject line so it is more specific than "Re: Contents of users digest..."

Today's Topics:

   1. Re: Issue with reporting variables on SoGE 8.1.8 - variables
      and their values dissapear after some time (William Hay)


Message: 1
Date: Tue, 24 Mar 2015 12:08:00 +0000
From: William Hay <w.hay at ucl.ac.uk>
To: users at gridengine.org
Subject: Re: [gridengine users] Issue with reporting variables on SoGE
	8.1.8 - variables and their values dissapear after some time
Message-ID: <20150324120800.abd44a9c662e309c14789a75 at ucl.ac.uk>
Content-Type: text/plain; charset="utf-8"

On Tue, 24 Mar 2015 11:26:04 +0000
Yuri Burmachenko <yuribu at mellanox.com> wrote:

> Hallo to distinguished forum members,
> I hope you can assist me.
> We are in process of transition to SoGE 8.1.8 from SGE GE 6.1u6.
> We experience strange issues with reporting variables.
> For example we have a reporting variable called mt_os_type which returns OS type.
> Output example 1: centos6
> Output example 1: centos5
> #name                  shortcut               type        relop requestable consumable default  urgency
> #-------------------------------------------------------------------------------------------------------
> mt_os_type             mt_os_type             STRING      ==    YES         NO         NONE     0
> qconf -se global | grep -i mt_os_typeL
>                       mt_syslog_bytes,mt_yp_query_ms,mt_os_type
> It works OK on SGE GE 6.1.u6, while on SoGE 8.1.8 it behaves inconsistently: it appears and after some time the variable and its value disappears from attributes on execution hosts.
> qconf ?se <some_host> | /bin/grep ?i <some_host>:
> mt_os_type=centos5
> And after a minute:
> qconf ?se <some_host> | /bin/grep ?i <some_host>:
> No variable and no value?
> Some additional information:
> Report time is (qconf ?sconf):
> load_report_time             00:00:40
> Within load_sensors.sh script we use several types of sensors: the ones which are invoked every 40 sec and other that invoked every 2 and 5 minutes, see below:

This looks like working as designed to me.  A load sensor value hasn't been reported for multiples of the load_report_time so it is now unknown.  If this didn't happen a problem that broke the load sensor wouldn't be detected.
If you are trying to avoid running load sensors more than necessary I would suggest caching the results of the infrequently run sensors and reporting the cached values every time.  You can save the output of the infrequent sensors to a file and then  cat it back every time the load_sensors.sh script is prompted for values.

William Hay <w.hay at ucl.ac.uk>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://gridengine.org/pipermail/users/attachments/20150324/6d8acbba/attachment-0001.sig>


users mailing list
users at gridengine.org

End of users Digest, Vol 51, Issue 10

More information about the users mailing list