LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Newbie (http://www.linuxquestions.org/questions/linux-newbie-8/)
-   -   Red Hat Cluster Issue? (http://www.linuxquestions.org/questions/linux-newbie-8/red-hat-cluster-issue-662745/)

your_shadow03 08-14-2008 09:21 AM

Red Hat Cluster Issue?
 
Guys...

I have two Cluster Nodes: Edwin1 and Edwin2.
Edwin2 is NFS Server while Edwin1 is NFS Client.
Anyway,NFS will be running on Active/Active ie both are running NFS simultanously.

On Edwin2 we can see the following configuration:

# df –h
10.227.167.5:/usr/bang-test/xml on /var/tmp/kunal type nfs (rw,soft,addr=10.227.167.5)
/dev/sdj1 on /usr/bang-test type ext3 (rw)

[[[[Note: As for Now /dev/sdj1 which is a shared storage partition is right now mounted on
Edwin2.As We do Clusvcadm -r bang -m edwin1 command,it will relocate to edwin1 node.]]]

I have mounted /usr/bang-test/xml to /var/tmp/kunal as seen above.
Corresponding to that I made the entry under NFS Mount on LUCI as:

NFS Mount Resource Configuration

Name ---> NFSMount
Mount point ---> /var/tmp/kunal
Host --> 10.227.169.3
Export path --> 10.227.167.5:/usr/bang-test/xml
NFS version NFS3 -> <Default>
NFS4
Options

OK.


On Edwin1 Now, The mount point /dev/sdj1 couldnt be seen.(After failover it will be seen)

#df -h

10.227.167.5:/usr/bang-test/xml
206G 360M 195G 1% /var/tmp/kunal

Now When I have written another Service Script called bang which mean now I have two Script
in hand.I have added both to the cluster.

As you see output of Clustat:

#clustat

edwin2# clustat
Member Status: Quorate

Member Name Status
------ ---- ------
edwin1-cluster Online, rgmanager
edwin2-cluster Online, Local, rgmanager

Service Name Owner (Last) State
------- ---- ----- ------ -----
bang edwin2-cluster started
NFSCluster edwin1-cluster started

So Now Both the service are added.When I now perform failover it doesnt appear working.
Is there anything mischievious I am doing?

Pls Advise

brianmcgee 08-16-2008 04:11 AM

Scripts you are using?
/etc/cluster/cluster.conf?

your_shadow03 08-18-2008 01:31 AM

Ya...my cluster.conf is:
Code:

<?xml version="1.0"?>
<cluster config_version="47" name="NGP-Cluster">
        <cman expected_votes="1" two_node="1"/>
        <clusternodes>
                <clusternode name="edwin1-cluster" nodeid="1" votes="1">
                        <fence>
                                <method name="hardware">
                                        <device ipaddr="edwin1-ilo" name="ipmilan"/>
                                </method>
                                <method name="last_resort">
                                        <device ipaddr="edwin1-cluster" name="last_resort"/>
                                </method>
                        </fence>
                </clusternode>
                <clusternode name="edwin2-cluster" nodeid="2" votes="1">
                        <fence>
                                <method name="hardware">
                                        <device ipaddr="edwin2-ilo" name="ipmilan"/>
                                </method>
                                <method name="last_resort">
                                        <device ipaddr="edwin2-cluster" name="last_resort"/>
                                </method>
                        </fence>
                </clusternode>
        </clusternodes>
        <fencedevices>
                <fencedevice agent="fence_ipmilan" name="ipmilan" option="off" passwd="@#$@#"/>
                <fencedevice agent="fence_manual" name="last_resort"/>
        </fencedevices>
        <rm log_facility="local3" log_level="7">
                <failoverdomains>
                        <failoverdomain name="PreferredFirst" ordered="1" restricted="1">
                                <failoverdomainnode name="edwin1-cluster" priority="0"/>
                                <failoverdomainnode name="edwin2-cluster" priority="1"/>
                        </failoverdomain>
                </failoverdomains>
                <resources/>
                <service autostart="1" domain="PreferredFirst" name="bang" recovery="relocate">
                        <script file="/home/fsadmin/featureserver/scripts/featureserver.sh" name="bang"/>
                        <ip address="10.227.167.5" monitor_link="0"/>
                        <fs device="/dev/sdj1" force_unmount="1" fstype="ext3" mountpoint="/usr/bang-test" name="/usr/bang-test"/>
                </service>
  <service autostart="1" domain="PreferredFirst" exclusive="0" name="NFSCluster" recovery="relocate">
                        <netfs exportpath="10.227.167.5:/usr/bang-test/xml" force_unmount="0" host="10.227.167.3" mountpoint="/var/tmp/kunal" name="NFSMount" nfstype="nfs"/>
                </service>
        </rm>
        <quorumd device="/dev/sdc1" interval="1" label="priquorum" log_facility="local3" log_level="7" tko="10" votes="2"/>
        <fence_daemon clean_start="1" post_fail_delay="0" post_join_delay="3"/>
</cluster>


brianmcgee 08-19-2008 02:11 AM

So you expect to mount the /dev/sdj1 on both nodes simultanously, but you only see that it gets mounted, if you failover the service?

If so, thats normal behaviour. The device is formatted with ext3, which may not be mounted on both nodes. In your case the ext3 filesystem failovers with the service.

You use a shared storage device, you say. That's the first step. The next step involves formatting the device with a clustered filesystem. In this case you would use Red Hat's Global Filesystem. Then you may mount the filesystem simultanously on both nodes manual or via /etc/fstab.


All times are GMT -5. The time now is 07:24 PM.