Usage of HPC and HA at BITS Pilani - Revision history

Biswajit at 14:19, 29 December 2012

2012-12-29T14:19:22Z

← Older revision		Revision as of 14:19, 29 December 2012
Line 1:		Line 1:
	[[category:Bits pilani]		[[category:Bits pilani]]
	=='''USAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''==		=='''USAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''==

Line 26:		Line 26:
	<nowiki># clustat</nowiki>		<nowiki># clustat</nowiki>

	[[Image:bits-3.png]]		[[Image:bits-2.png]]

	'''Stop manually HA Cluster by:'''		'''Stop manually HA Cluster by:'''

Biswajit at 14:18, 29 December 2012

2012-12-29T14:18:29Z

← Older revision		Revision as of 14:18, 29 December 2012
Line 1:		Line 1:
	~~<center>~~'''~~UGAGE~~ DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''~~</center>~~		[[category:Bits pilani]
			=='''USAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''==

			'''NAVDEEP MATHUR'''

	~~<center>'''NAVDEEP MATHUR'''</center>~~

			=='''HIGH AVAILABILITY CLUSTER'''==
	~~<center>~~'''HIGH AVAILABILITY CLUSTER'''~~</center>~~

	Two Nodes on HA:		Two Nodes on HA:
Line 16:		Line 16:
	'''How to Fence from one cluster to another manually , Run command from any node'''		'''How to Fence from one cluster to another manually , Run command from any node'''

	<nowiki># clusvcadm –d nfs</nowiki>		<nowiki># clusvcadm –d nfs</nowiki>

	Here nfs is an cluster service		Here nfs is an cluster service

	<nowiki># clusvcadm –e nfs –m `hostname`</nowiki>		<nowiki># clusvcadm –e nfs –m `hostname`</nowiki>

	How to check status of cluster		How to check status of cluster

	<nowiki># clustat</nowiki>		<nowiki># clustat</nowiki>

	[[Image:]]		[[Image:bits-3.png]]

	'''Stop manually HA Cluster by:'''		'''Stop manually HA Cluster by:'''

	<nowiki># service cman stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>		<nowiki># service cman stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>

	<nowiki># service rgmanager stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>		<nowiki># service rgmanager stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>

	'''Start manually HA Cluster by:'''		'''Start manually HA Cluster by:'''

	<nowiki># service cman start //Simultaneously as root on ibm-39 and ibm-40</nowiki>		<nowiki># service cman start //Simultaneously as root on ibm-39 and ibm-40</nowiki>

	<nowiki># service rgmanager start //Simultaneously as root on ibm-39 and ibm-40</nowiki>		<nowiki># service rgmanager start //Simultaneously as root on ibm-39 and ibm-40</nowiki>

	Then check the status of cluster by “clustat” command and also ping to 172.20.4.43 that is your virtual ip of cluster.		Then check the status of cluster by “clustat” command and also ping to 172.20.4.43 that is your virtual ip of cluster.
Line 52:		Line 52:
	'''On Node2 :'''		'''On Node2 :'''

	<nowiki># service rgmanager stop</nowiki>		<nowiki># service rgmanager stop</nowiki>
			<nowiki># service cman stop</nowiki>
	<nowiki># service cman stop</nowiki>

	Then Simultaneously on '''Node1 and Node2'''		Then Simultaneously on '''Node1 and Node2'''

	<nowiki># service cman start</nowiki>		<nowiki># service cman start</nowiki>

	<nowiki># service rgmanager start</nowiki>		<nowiki># service rgmanager start</nowiki>

	And check clustat		And check

			clustat


	~~<center>~~'''HIGH PROCESSING CLUSTER'''~~</center>~~		=='''HIGH PROCESSING CLUSTER'''==

	'''When System restart run command as wolf and lam user'''		'''When System restart run command as wolf and lam user'''

	~~<center>~~'''For HPC1 -2 node HPC'''~~</center>~~		'''For HPC1 -2 node HPC'''

	'''Login as wolf'''		'''Login as wolf'''

	<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>		<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>
			<nowiki>#mpdtrace </nowiki>// check active nodes
	<nowiki>#mpdtrace </nowiki>// check active nodes

	This will active mpd deamon on all the nodes		This will active mpd deamon on all the nodes
Line 81:		Line 80:
	'''Login as lam'''		'''Login as lam'''

	<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>		<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>
			<nowiki>#mpdtrace // check active nodes</nowiki>
	<nowiki>#mpdtrace // check active nodes</nowiki>

	This will active mpd deamon on all the nodes		This will active mpd deamon on all the nodes


	~~<center>~~'''For HPC2 -9 node HPC'''~~</center>~~		'''For HPC2 -9 node HPC'''

	'''Login as wolf2'''		'''Login as wolf2'''

	<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>		<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>
			<nowiki>#mpdtrace // check active nodes</nowiki>
	<nowiki>#mpdtrace // check active nodes</nowiki>

	This will active mpd deamon on all the nodes		This will active mpd deamon on all the nodes
Line 100:		Line 97:
	'''Login as lam1'''		'''Login as lam1'''

	<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>		<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>
			<nowiki>#mpdtrace //check active nodes</nowiki>
	<nowiki>#mpdtrace //check active nodes</nowiki>

	This will active mpd deamon on all the nodes		This will active mpd deamon on all the nodes

	Then mpi commands execute using mpiexec command		Then mpi commands execute using mpiexec command

Biswajit: Created page with "'''UGAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER''' '''NAVDEEP MATHUR''' '''HIGH AVAILABILITY CLUSTER''' ..."

2012-12-29T14:12:58Z

Created page with "<center>'''UGAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''</center> <center>'''NAVDEEP MATHUR'''</center> <center>'''HIGH AVAILABILITY CLUSTER'''</center> ..."

New page

<center>'''UGAGE DOCUMENT OF HPC CLUSTER AND HIGH AVAILABILITY CLUSTER'''</center>

<center>'''NAVDEEP MATHUR'''</center>

<center>'''HIGH AVAILABILITY CLUSTER'''</center>

Two Nodes on HA:

# ibm-39.bits.ac – 172.20.4.39
# ibm-40.bits.ac – 172.20.4.40

Virtual IP of HA : 172.20.4.43

'''How to Fence from one cluster to another manually , Run command from any node'''

<nowiki># clusvcadm –d nfs</nowiki>

Here nfs is an cluster service

<nowiki># clusvcadm –e nfs –m `hostname`</nowiki>

How to check status of cluster

<nowiki># clustat</nowiki>

[[Image:]]

'''Stop manually HA Cluster by:'''

<nowiki># service cman stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>

<nowiki># service rgmanager stop //Simultaneously as root on ibm-39 and ibm-40</nowiki>

'''Start manually HA Cluster by:'''

<nowiki># service cman start //Simultaneously as root on ibm-39 and ibm-40</nowiki>

<nowiki># service rgmanager start //Simultaneously as root on ibm-39 and ibm-40</nowiki>

Then check the status of cluster by “clustat” command and also ping to 172.20.4.43 that is your virtual ip of cluster.

'''Note :'''

# When first server is out of the network then it fenced to second server and at same time first server has automatically forcefully rebooted
# Then virtual ip switch to second server.
# When first server has been UP then need to switch cluster manually.
# You have two choices by clusvcadm or restart rgmanager and cman.
# We take here second option i.e. restart rgmanager and cman

'''On Node2 :'''

<nowiki># service rgmanager stop</nowiki>

<nowiki># service cman stop</nowiki>

Then Simultaneously on '''Node1 and Node2'''

<nowiki># service cman start</nowiki>

<nowiki># service rgmanager start</nowiki>

And check clustat

<center>'''HIGH PROCESSING CLUSTER'''</center>

'''When System restart run command as wolf and lam user'''

<center>'''For HPC1 -2 node HPC'''</center>

'''Login as wolf'''

<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>

<nowiki>#mpdtrace </nowiki>// check active nodes

This will active mpd deamon on all the nodes

'''Login as lam'''

<nowiki>#mpdboot –n 2 –f mpd.hosts</nowiki>

<nowiki>#mpdtrace // check active nodes</nowiki>

This will active mpd deamon on all the nodes

<center>'''For HPC2 -9 node HPC'''</center>

'''Login as wolf2'''

<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>

<nowiki>#mpdtrace // check active nodes</nowiki>

This will active mpd deamon on all the nodes

'''Login as lam1'''

<nowiki>#mpdboot –n 9 –f mpd.hosts</nowiki>

<nowiki>#mpdtrace //check active nodes</nowiki>

This will active mpd deamon on all the nodes

Then mpi commands execute using mpiexec command