HelpPageForAllScripts - Revision history

Rougny Romain rrougny@ua.ac.be at 10:29, 27 January 2017

2017-01-27T10:29:29Z

← Older revision		Revision as of 10:29, 27 January 2017
Line 1:		Line 1:
	=== CE ===		== CE ==

	* '''manage_queues.py'''		* '''manage_queues.py'''
Line 44:		Line 44:


	=== QNAT ===		== QNAT ==

	* '''dns-manager'''		* '''dns-manager'''

Rougny Romain rrougny@ua.ac.be at 10:26, 27 January 2017

2017-01-27T10:26:29Z

← Older revision		Revision as of 10:26, 27 January 2017
Line 42:		Line 42:
	* '''torque-user-info.py'''		* '''torque-user-info.py'''
	Using same infra as '''parse_torque_accounting_log''', will print details of all jobs per user. More details [[Cluster_Presentation#Job_Statistics\|here]].		Using same infra as '''parse_torque_accounting_log''', will print details of all jobs per user. More details [[Cluster_Presentation#Job_Statistics\|here]].


			=== QNAT ===

			* '''dns-manager'''
			<pre>
			dns-manager ADD\|DELETE $HOST $IP

			example: ./dns-manager ADD schtroumpf.wn.iihe.ac.be 192.168.66.6
			</pre>

			* '''get-free-ip'''
			<pre>
			Usage:

			get-free-ip 192.168[.10][.65] [-d]

			IP :
			The ip provided must be in one of the ranges owned by IIHE.
			You can omit AT MOST the 2 last numbers of an ip.
			Giving a complete ip will return the first free one starting from the
			provided ip. It is useful to get IPs from subnets not /16 or /24, eg:
			get-free-ip 193.58.172.128 will give free IPs in 193.58.172.128/25


			-d:
			Enables debugging. Must be put after IP.
			All messages go to stderr, only the final ip is printed to stdout.
			[Only errors that exit the program print their messages to stdout too.]
			</pre>

			* '''reboot-wns'''
			<pre>
			Usage:
			reboot-wns --pre-reboot
			reboot-wns --post-reboot
			reboot-wns --server [--force-reboot] [--offline-slots <NSLOTS>]

			--server: start on qnat the server mode to reboot all WNs automatically
			Logfile stored in /tmp/reboot-wns.log
			OPTIONS:

			[--force-reboot] : will use special reboot forcing it, rather than reboot funtion.
			Useful when nfs is hanging

			[--offline-slots <NSLOTS>] : specify the amount of slots you want to have offline.
			Defaults to 700

			[--continue] : will lookup the status of the last time reboot-cluster --server was run
			and will continue from there. This allows to specify what still needs to
			reboot by editing the files:
			/tmp/listOfNodesStillToReboot
			/tmp/listOfNodesPutBackInProd


			--pre-reboot: executed on WN to check if reboot can be performed.
			Exit code of 1 if the test sequence fails

			--post-reboot: executed on WN to check if node can go back into prod
			Exit code of 1 if the test sequence fails
			</pre>

Rougny Romain rrougny@ua.ac.be: Created page with "=== CE === * '''manage_queues.py'''

 Usage: manage_queues.py [OPTION] Manage Torque queues    -h,  --help			Show this help   -s,  --status			Show queues status        --..."

2017-01-27T10:16:16Z

Created page with "=== CE === * '''manage_queues.py''' <pre> Usage: manage_queues.py [OPTION] Manage Torque queues -h, --help Show this help -s, --status Show queues status --..."

New page

=== CE ===

* '''manage_queues.py'''
<pre>
Usage: manage_queues.py [OPTION]
Manage Torque queues

-h, --help Show this help
-s, --status Show queues status
--close-all Close all opened queues
--open-all Open all closed queues
--show-all Show all attributes for all queues
--close QUEUE_NAME Close queue QUEUE_NAME
--open QUEUE_NAME Open queue QUEUE_NAME
--show QUEUE_NAME Show all attributes for queue QUEUE_NAME
</pre>

* '''myce.sh'''
<pre>
Valid args :
- list-running-nodes ( -r )
- list-up-nodes ( -u )
- list-down-nodes ( -d )
- count-cpu (-c) [queue]
</pre>

* '''Releasehold'''
releaseholds jobs. For all, just do: ./Releasehold Deferred

* '''Requeue_jobs'''
No argument !! Need 2 arguments to grep on

* '''Restart-creamce'''
Will restart the creamce

* '''Restart_pbs_maui'''
Will restart maui & pbs in the right sequence

* '''parse_torque_accounting_log.py'''
Will show summary info of all jobs for the last 8 days

* '''torque-user-info.py'''
Using same infra as '''parse_torque_accounting_log''', will print details of all jobs per user. More details [[Cluster_Presentation#Job_Statistics|here]].