<div dir="ltr"><div><div>Thanks, Soumya, for giving me a confirmation on the time for the take-over process. That's very helpful.</div></div><div><br></div><div>The 'showmount' got hung on the destination host while i see ganesha-nfsd was running. </div><div>On that, 1 thing i tried out-of-ordinary was to set a 2nd set of *cluster_ip for the storage-nodes using 'pcs'. Meaning, other than the VIP i have in 'ganesha-ha.conf', i also used 'pcs' to create another set of VIPs, and thought of using 'constraint colocation' to get both set of VIPs and NFSd going. Though both set of VIPs did failover to the destination node, but the ganesha-nfsd was hung at the destination node when the source node got brought down. I am guessing this setup probably caused the confusion of nfs-ganesha ... ? Is that a No-No to have more than 1 set of VIPs for NFS-Ganesha in different subnets?</div><div><br></div><div>Any suggestions, and advice will be appreciated.</div><div><br></div><div><br></div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Sun, Feb 5, 2017 at 11:50 PM, Soumya Koduri <span dir="ltr"><<a href="mailto:skoduri@redhat.com" target="_blank">skoduri@redhat.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class=""><br>
<br>
On 02/04/2017 06:20 AM, ML Wong wrote:<br>
</span><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">
Thanks so much for your promptly response, Soumya.<br>
That helps clearing out one of my questions. I am trying to figure out<br>
why NFS service did not failover/pick-up the NFS clients last time when<br>
one of our cluster-nodes failed.<br>
<br>
Though i could see, in corosync.log, a notify got sent to the cluster<br>
the failed node, the election, and the IP failover process seems to all<br>
be finished with in around minute. However, after the IP failover to the<br>
destinated node, i tried to do a "showmount -e localhost" - the command<br></span>
got hung.But, i still see ganesha-nfsd is running in the host.<br>
</blockquote>
<br>
Is it on destination node? I mean does 'showmount -e localhost' get hung on destination node post IP failover.<span class=""><br>
<br>
To your<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
expertise, if i understand the process correctly, given that i keep all<br>
the default timeout/interval settings for nfs-mon, nfs-grace, the entire<br>
IP failover, and NFS service failover process should be completed within<br>
2 minutes. Am i correct?<br>
</blockquote>
<br></span>
Thats right.<br>
<br>
Thanks,<br>
Soumya<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">
<br>
Your help is again appreciated.<br>
<br>
On Thu, Feb 2, 2017 at 11:42 PM, Soumya Koduri <<a href="mailto:skoduri@redhat.com" target="_blank">skoduri@redhat.com</a><br></span><div><div class="h5">
<mailto:<a href="mailto:skoduri@redhat.com" target="_blank">skoduri@redhat.com</a>>> wrote:<br>
<br>
Hi,<br>
<br>
On 02/03/2017 07:52 AM, ML Wong wrote:<br>
<br>
Hello All,<br>
Any pointers will be very-much appreciated. Thanks in advance!<br>
<br>
Environment:<br>
Running centOS 7.2.511<br>
Gluster: 3.7.16, with nfs-ganesha on 2.3.0.1 from<br>
centos-gluster37 repo<br>
sha1: cab5df4064e3a31d1d92786d91bd41<wbr>d91517fba8 ganesha-ha.sh<br>
<br>
we have used this set up in 3 different gluster, nfs-ganesha<br>
environment. The cluster got setup when we do 'gluster nfs-ganesha<br>
enable' , and we can serve NFS without issues. And i see all the<br>
resources got created, but not the *hostname*-trigger_ip-1<br>
resources? Is<br>
that normal?<br>
<br>
<br>
Yes it is normal. With change [1], new resource agent attributes<br>
have been introduced in place of *-trigger_ip-1 to monitor, move the<br>
VIP and put the cluster in grace. More details are in the change#<br>
commit msg.<br>
<br>
Thanks,<br>
Soumya<br>
<br>
[1]<br>
<a href="https://github.com/gluster/glusterfs/commit/e8121c4afb3680f532b450872b5a3ffcb3766a97" rel="noreferrer" target="_blank">https://github.com/gluster/glu<wbr>sterfs/commit/e8121c4afb3680f5<wbr>32b450872b5a3ffcb3766a97</a><br>
<<a href="https://github.com/gluster/glusterfs/commit/e8121c4afb3680f532b450872b5a3ffcb3766a97" rel="noreferrer" target="_blank">https://github.com/gluster/gl<wbr>usterfs/commit/e8121c4afb3680f<wbr>532b450872b5a3ffcb3766a97</a>><br>
<br>
without *hostname*-trigger_ip-1, according to ganesha-ha.sh,<br>
wouldn't it<br>
affect the NFS going into grace, and help to transition the NFS<br>
service<br>
to other member nodes at the times of node-failures? please<br>
correct me<br>
if i misunderstood.<br>
<br>
I tried issuing both 'gluster nfs-ganesha enable', and 'bash -x<br>
/usr/libexec/ganesha/ganesha-h<wbr>a.sh --setup'. In both scenarios,<br>
i still<br>
don't see the *hostname*-trigger_ip-1 got created.<br>
<br>
below is my ganesha-ha.conf<br>
HA_NAME="ganesha-ha-01"<br>
HA_VOL_SERVER="vm-fusion1"<br>
HA_CLUSTER_NODES="vm-fusion1,v<wbr>m-fusion3"<br>
VIP_vm-fusion1="192.168.30.211<wbr>"<br>
VIP_vm-fusion3="192.168.30.213<wbr>"<br>
<br>
<br>
<br>
<br>
______________________________<wbr>_________________<br>
Gluster-users mailing list<br></div></div>
<a href="mailto:Gluster-users@gluster.org" target="_blank">Gluster-users@gluster.org</a> <mailto:<a href="mailto:Gluster-users@gluster.org" target="_blank">Gluster-users@gluster.<wbr>org</a>><br>
<a href="http://lists.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://lists.gluster.org/mailm<wbr>an/listinfo/gluster-users</a><br>
<<a href="http://lists.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://lists.gluster.org/mail<wbr>man/listinfo/gluster-users</a>><br>
<br>
<br>
</blockquote>
</blockquote></div><br></div>