Login | Register For Free | Help
Search for: (Advanced)

Mailing List Archive: Linux-HA: Users

Running multiple Oracle Database Instances on active/active Linux-HA cluster

 

 

Linux-HA users RSS feed   Index | Next | Previous | View Threaded


meryem.belhadj at gmail

Nov 3, 2009, 9:05 AM

Post #1 of 5 (220 views)
Permalink
Running multiple Oracle Database Instances on active/active Linux-HA cluster

Hello,

We have a two nodes active/active heartbeat cluster. We configured two
resource groups containing all necessary volume group, file system, ip
address, database instance and listener resources. We run a resource group
on each node. Heartbeat package version is: heartbeat-2.1.3-0.9 installed on
SLES 10 SP2 ppc.
The issue we are facing is when we try to run the two resource groups on one
single node simulating a failover. In fact, when a node (node 1 for example)
fails, the resource group that it owns is migrated to the second node (node
2 for example) but we noticed that the database instance that is running on
this node (node 2) is retarted! Which means that we have both database
instances retarted.

Is that normal? And how avoid that running database is restarted?
We tried two configuration methods using OCF and LSB Oracle resources but we
face the same problem.

We can attach logs or configuration files if needed.

Please help us.

Best regards.

Meryem Belhadj.
System Engineer.
_______________________________________________
Linux-HA mailing list
Linux-HA[at]lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems


wgumularz at gmail

Nov 3, 2009, 9:53 AM

Post #2 of 5 (207 views)
Permalink
Re: Running multiple Oracle Database Instances on active/active Linux-HA cluster [In reply to]

2009/11/3 Meryem Belhadj <meryem.belhadj[at]gmail.com>

> Hello,
>
> We have a two nodes active/active heartbeat cluster. We configured two
> resource groups containing all necessary volume group, file system, ip
> address, database instance and listener resources. We run a resource group
> on each node. Heartbeat package version is: heartbeat-2.1.3-0.9 installed
> on
> SLES 10 SP2 ppc.
> The issue we are facing is when we try to run the two resource groups on
> one
> single node simulating a failover. In fact, when a node (node 1 for
> example)
> fails, the resource group that it owns is migrated to the second node (node
> 2 for example) but we noticed that the database instance that is running on
> this node (node 2) is retarted! Which means that we have both database
> instances retarted.
>

Your problem seems to be related to this topic:
http://lists.linux-ha.org/pipermail/linux-ha/2008-December/035684.html

Regards,
Wojciech Gumularz

>
> Is that normal? And how avoid that running database is restarted?
> We tried two configuration methods using OCF and LSB Oracle resources but
> we
> face the same problem.
>
> We can attach logs or configuration files if needed.
>
> Please help us.
>
> Best regards.
>
> Meryem Belhadj.
> System Engineer.
> _______________________________________________
> Linux-HA mailing list
> Linux-HA[at]lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
_______________________________________________
Linux-HA mailing list
Linux-HA[at]lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems


meryem.belhadj at gmail

Nov 4, 2009, 7:18 AM

Post #3 of 5 (196 views)
Permalink
Re: Running multiple Oracle Database Instances on active/active Linux-HA cluster [In reply to]

Hello,

Thanks for the reply.

I had a look to the link you provided but it's not the same issue because in
our configuration we use resource groups ans don't use clones which is quite
different.

I attached our cib.xml file.

Best regards,

Meryem Belhadj.
System Engineer.

2009/11/3 Wojciech Gumularz <wgumularz[at]gmail.com>

> 2009/11/3 Meryem Belhadj <meryem.belhadj[at]gmail.com>
>
> > Hello,
> >
> > We have a two nodes active/active heartbeat cluster. We configured two
> > resource groups containing all necessary volume group, file system, ip
> > address, database instance and listener resources. We run a resource
> group
> > on each node. Heartbeat package version is: heartbeat-2.1.3-0.9 installed
> > on
> > SLES 10 SP2 ppc.
> > The issue we are facing is when we try to run the two resource groups on
> > one
> > single node simulating a failover. In fact, when a node (node 1 for
> > example)
> > fails, the resource group that it owns is migrated to the second node
> (node
> > 2 for example) but we noticed that the database instance that is running
> on
> > this node (node 2) is retarted! Which means that we have both database
> > instances retarted.
> >
>
> Your problem seems to be related to this topic:
> http://lists.linux-ha.org/pipermail/linux-ha/2008-December/035684.html
>
> Regards,
> Wojciech Gumularz
>
> >
> > Is that normal? And how avoid that running database is restarted?
> > We tried two configuration methods using OCF and LSB Oracle resources but
> > we
> > face the same problem.
> >
> > We can attach logs or configuration files if needed.
> >
> > Please help us.
> >
> > Best regards.
> >
> > Meryem Belhadj.
> > System Engineer.
> > _______________________________________________
> > Linux-HA mailing list
> > Linux-HA[at]lists.linux-ha.org
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
> >
> _______________________________________________
> Linux-HA mailing list
> Linux-HA[at]lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
Attachments: cib.xml (9.42 KB)


dejanmm at fastmail

Nov 4, 2009, 7:30 AM

Post #4 of 5 (196 views)
Permalink
Re: Running multiple Oracle Database Instances on active/active Linux-HA cluster [In reply to]

Hi,

On Wed, Nov 04, 2009 at 04:18:28PM +0100, Meryem Belhadj wrote:
> Hello,
>
> Thanks for the reply.
>
> I had a look to the link you provided but it's not the same issue because in
> our configuration we use resource groups ans don't use clones which is quite
> different.

I'm really not sure, but I'd say that the issue is not
necessarily related to clones.

If you think that this is a bug, please open a bugzilla with
hb_report. Though I'm not sure if your version (2.1.x) is going
to get necessary attention.

Thanks,

Dejan

> I attached our cib.xml file.
>
> Best regards,
>
> Meryem Belhadj.
> System Engineer.
>
> 2009/11/3 Wojciech Gumularz <wgumularz[at]gmail.com>
>
> > 2009/11/3 Meryem Belhadj <meryem.belhadj[at]gmail.com>
> >
> > > Hello,
> > >
> > > We have a two nodes active/active heartbeat cluster. We configured two
> > > resource groups containing all necessary volume group, file system, ip
> > > address, database instance and listener resources. We run a resource
> > group
> > > on each node. Heartbeat package version is: heartbeat-2.1.3-0.9 installed
> > > on
> > > SLES 10 SP2 ppc.
> > > The issue we are facing is when we try to run the two resource groups on
> > > one
> > > single node simulating a failover. In fact, when a node (node 1 for
> > > example)
> > > fails, the resource group that it owns is migrated to the second node
> > (node
> > > 2 for example) but we noticed that the database instance that is running
> > on
> > > this node (node 2) is retarted! Which means that we have both database
> > > instances retarted.
> > >
> >
> > Your problem seems to be related to this topic:
> > http://lists.linux-ha.org/pipermail/linux-ha/2008-December/035684.html
> >
> > Regards,
> > Wojciech Gumularz
> >
> > >
> > > Is that normal? And how avoid that running database is restarted?
> > > We tried two configuration methods using OCF and LSB Oracle resources but
> > > we
> > > face the same problem.
> > >
> > > We can attach logs or configuration files if needed.
> > >
> > > Please help us.
> > >
> > > Best regards.
> > >
> > > Meryem Belhadj.
> > > System Engineer.
> > > _______________________________________________
> > > Linux-HA mailing list
> > > Linux-HA[at]lists.linux-ha.org
> > > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > > See also: http://linux-ha.org/ReportingProblems
> > >
> > _______________________________________________
> > Linux-HA mailing list
> > Linux-HA[at]lists.linux-ha.org
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
> >

> <cib generated="true" admin_epoch="0" epoch="10" num_updates="1" have_quorum="true" ignore_dtd="false" num_peers="2" ccm_transition="2" cib_feature_revision="2.0" crm_feature_set="2.0" dc_uuid="e3b0cf04-e13c-43d5-9cff-3d7b1aac8538" cib-last-written="Wed Sep 23 16:11:14 2009">
> <configuration>
> <crm_config>
> <cluster_property_set id="cib-bootstrap-options">
> <attributes>
> <nvpair id="cib-bootstrap-options-dc-version" name="dc-version" value="2.1.3-node: a3184d5240c6e7032aef9cce6e5b7752ded544b3"/>
> </attributes>
> </cluster_property_set>
> </crm_config>
> <nodes>
> <node id="581661ec-ce3e-4a75-b29f-6a12dbc8f03e" uname="ttndb1" type="normal"/>
> <node id="e3b0cf04-e13c-43d5-9cff-3d7b1aac8538" uname="ttndb2" type="normal"/>
> </nodes>
> <resources>
> <group id="liasse-srv" ordered="true" collocated="true">
> <primitive id="liasse-vg" class="ocf" type="LVM" provider="heartbeat">
> <instance_attributes id="liasse-vg_instance_attrs">
> <attributes>
> <nvpair id="liasse-vg-volgrpname" name="volgrpname" value="liasse-vg"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="liasse-vg-monitor" name="monitor" interval="10" timeout="30" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="liasse-fs" class="ocf" type="Filesystem" provider="heartbeat">
> <instance_attributes id="liasse-fs_instance_attrs">
> <attributes>
> <nvpair id="liasse-fs-fstype" name="fstype" value="reiserfs"/>
> <nvpair id="liasse-fs-device" name="device" value="/dev/liasse-vg/liasse-lv"/>
> <nvpair id="liasse-fs-directory" name="directory" value="/liasse"/>
> <nvpair id="liasse-fs-options" name="options" value="acl,user_xattr"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="liasse-fs-monitor" name="monitor" interval="20" timeout="40" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="fb-liasse-vg" class="ocf" type="LVM" provider="heartbeat">
> <instance_attributes id="fb-liasse-vg_instance_attrs">
> <attributes>
> <nvpair id="fb-liasse-vg-volgrpname" name="volgrpname" value="fb-liasse-vg"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="fb-liasse-vg-monitor" name="monitor" interval="10" timeout="30" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="fb-liasse-fs" class="ocf" type="Filesystem" provider="heartbeat">
> <instance_attributes id="fb-liasse-fs_instance_attrs">
> <attributes>
> <nvpair id="fb-liasse-fs-fstype" name="fstype" value="reiserfs"/>
> <nvpair id="fb-liasse-fs-device" name="device" value="/dev/fb-liasse-vg/fb-liasse-lv"/>
> <nvpair id="fb-liasse-fs-directory" name="directory" value="/flashback_liasse"/>
> <nvpair id="fb-liasse-fs-options" name="options" value="acl,user_xattr"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="fb-liasse-fs-monitor" name="monitor" interval="20" timeout="40" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="liasse-ip" class="ocf" type="IPaddr2" provider="heartbeat">
> <instance_attributes id="liasse-ip_instance_attrs">
> <attributes>
> <nvpair id="liasse-ip-ip" name="ip" value="10.150.3.8"/>
> <nvpair id="liasse-ip-nic" name="nic" value="bond0"/>
> <nvpair id="liasse-ip-cidr_netmask" name="cidr_netmask" value="255.255.255.0"/>
> <nvpair id="liasse-ip-iflabel" name="iflabel" value="liasse"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="liasse-ip-monitor" name="monitor" interval="10s" timeout="20s" start_delay="5s"/>
> </operations>
> </primitive>
> <primitive id="liasse-db" class="lsb" type="liassedb" provider="heartbeat"/>
> <primitive id="liasse-lsnr" class="lsb" type="liasselsnr" provider="heartbeat"/>
> <instance_attributes id="liasse-srv_instance_attrs">
> <attributes/>
> </instance_attributes>
> </group>
> <group id="web-srv" ordered="true" collocated="true">
> <primitive id="web-vg" class="ocf" type="LVM" provider="heartbeat">
> <instance_attributes id="web-vg_instance_attrs">
> <attributes>
> <nvpair id="web-vg-volgrpname" name="volgrpname" value="web-vg"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="web-vg-monitor" name="monitor" interval="10" timeout="30" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="web-fs" class="ocf" type="Filesystem" provider="heartbeat">
> <instance_attributes id="web-fs_instance_attrs">
> <attributes>
> <nvpair id="web-fs-fstype" name="fstype" value="reiserfs"/>
> <nvpair id="web-fs-device" name="device" value="/dev/web-vg/web-lv"/>
> <nvpair id="web-fs-directory" name="directory" value="/web"/>
> <nvpair id="web-fs-options" name="options" value="acl,user_xattr"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="web-fs-monitor" name="monitor" interval="20" timeout="40" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="fb-web-vg" class="ocf" type="LVM" provider="heartbeat">
> <instance_attributes id="fb-web-vg_instance_attrs">
> <attributes>
> <nvpair id="fb-web-vg-volgrpname" name="volgrpname" value="fb-web-vg"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="fb-web-vg-monitor" name="monitor" interval="10" timeout="30" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="fb-web-fs" class="ocf" type="Filesystem" provider="heartbeat">
> <instance_attributes id="fb-web-fs_instance_attrs">
> <attributes>
> <nvpair id="fb-web-fs-fstype" name="fstype" value="reiserfs"/>
> <nvpair id="fb-web-fs-device" name="device" value="/dev/fb-web-vg/fb-web-lv"/>
> <nvpair id="fb-web-fs-directory" name="directory" value="/flashback_web"/>
> <nvpair id="fb-web-fs-options" name="options" value="acl,user_xattr"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="fb-web-fs-monitor" name="monitor" interval="20" timeout="40" start_delay="10"/>
> </operations>
> </primitive>
> <primitive id="web-ip" class="ocf" type="IPaddr2" provider="heartbeat">
> <instance_attributes id="web-ip_instance_attrs">
> <attributes>
> <nvpair id="web-ip-ip" name="ip" value="10.150.3.9"/>
> <nvpair id="web-ip-nic" name="nic" value="bond0"/>
> <nvpair id="web-ip-cidr_netmask" name="cidr_netmask" value="255.255.255.0"/>
> <nvpair id="web-ip-iflabel" name="iflabel" value="web"/>
> </attributes>
> </instance_attributes>
> <operations>
> <op id="web-ip-monitor" name="monitor" interval="10s" timeout="20s" start_delay="5s"/>
> </operations>
> </primitive>
> <primitive id="web-db" class="lsb" type="webdb" provider="heartbeat"/>
> <primitive id="web-lsnr" class="lsb" type="weblsnr" provider="heartbeat"/>
> <instance_attributes id="web-srv_instance_attrs">
> <attributes/>
> </instance_attributes>
> </group>
> </resources>
> <constraints>
> <rsc_location id="liasse-srv-prefer-node1" rsc="liasse-srv">
> <rule id="liasse-srv-prefer-node1-rule" score="1000">
> <expression id="liasse-srv-prefer-node1-rule-exp" attribute="#uname" operation="eq" value="ttndb1"/>
> </rule>
> </rsc_location>
> <rsc_location id="liasse-srv-connected" rsc="liasse-srv">
> <rule id="liasse-srv-IS-connected" score="-INFINITY" boolean_op="or">
> <expression id="liasse-srv-IS-connected-not_defined" attribute="pingd" operation="not_defined"/>
> <expression id="liasse-srv-IS-connected-lte-0" attribute="pingd" operation="lte" value="0"/>
> </rule>
> </rsc_location>
> <rsc_location id="web-srv-prefer-node2" rsc="web-srv">
> <rule id="web-srv-prefer-node2-rule" score="1000">
> <expression id="web-srv-prefer-node2-rule-exp" attribute="#uname" operation="eq" value="ttndb2"/>
> </rule>
> </rsc_location>
> <rsc_location id="web-srv-connected" rsc="web-srv">
> <rule id="web-srv-IS-connected" score="-INFINITY" boolean_op="or">
> <expression id="web-srv-IS-connected-not_defined" attribute="pingd" operation="not_defined"/>
> <expression id="web-srv-IS-connected-lte-0" attribute="pingd" operation="lte" value="0"/>
> </rule>
> </rsc_location>
> <rsc_location id="cli-prefer-liasse-srv" rsc="liasse-srv">
> <rule id="prefered_cli-prefer-liasse-srv" score="INFINITY">
> <expression attribute="#uname" id="bee1cb05-7a7a-4ac1-a88f-758add54c335" operation="eq" value="ttndb2"/>
> </rule>
> </rsc_location>
> </constraints>
> </configuration>
> </cib>

> _______________________________________________
> Linux-HA mailing list
> Linux-HA[at]lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA[at]lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems


andrew at beekhof

Nov 5, 2009, 1:06 AM

Post #5 of 5 (189 views)
Permalink
Re: Running multiple Oracle Database Instances on active/active Linux-HA cluster [In reply to]

On Tue, Nov 3, 2009 at 6:05 PM, Meryem Belhadj <meryem.belhadj[at]gmail.com> wrote:
> Hello,
>
> We have a two nodes active/active heartbeat cluster. We configured two
> resource groups containing all necessary volume group, file system, ip
> address, database instance and listener resources. We run a resource group
> on each node. Heartbeat package version is: heartbeat-2.1.3-0.9 installed on
> SLES 10 SP2 ppc.

You should really get the latest updates from SUSE.
2.1.4 should be available by now IIRC and it had many fixes over 2.1.3

> The issue we are facing is when we try to run the two resource groups on one
> single node simulating a failover. In fact, when a node (node 1 for example)
> fails, the resource group that it owns is migrated to the second node (node
> 2 for example) but we noticed that the database instance that is running on
> this node (node 2) is retarted! Which means that we have both database
> instances retarted.
>
> Is that normal? And how avoid that running database is restarted?
> We tried two configuration methods using OCF and LSB Oracle resources but we
> face the same problem.
>
> We can attach logs or configuration files if needed.
>
> Please help us.
>
> Best regards.
>
> Meryem Belhadj.
> System Engineer.
> _______________________________________________
> Linux-HA mailing list
> Linux-HA[at]lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
_______________________________________________
Linux-HA mailing list
Linux-HA[at]lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Linux-HA users RSS feed   Index | Next | Previous | View Threaded
 
 


Interested in having your list archived? Contact lists@gossamer-threads.com
 
  Web Applications & Managed Hosting Powered by Gossamer Threads Inc.