[Ocfs2-users] 答复： Re: What do I need so ocfs2 logical volume mounted during boot

[Zhen Ren]

I don't think you have to get familiar with corosync/pacemaker, but pacemaker cluster stack can make things easier to get done.

Here is an example specific for SUSE linux.
      https://www.suse.com/documentation/sle_ha/book_sleha/data/cha_ha_ocfs2.html
It may show another way to do your task,but I have no idea how to do that on ubuntu.

--
Best regards,
Zhen, Ren
HA team, SUSE


>>> 
> so I put nc to check until port 7777 is up, then proceed with starting 
> ocfs2 & o2cb like below, which works 
>  
> while ! nc -z 10.224.202.192 7777 
> do 
>   sleep 5 
> done 
>  
> I'm not familiar with corosync/pacemaker. Do I need fencing device for 
> 2 nodes? Any configuration samples? 
>  
> 2015-03-17 13:14 GMT+08:00 Zhen Ren <zren at suse.com>: 
> > Yeah, it seems nodes in the same cluster cannot fully reach each other. 
> > This may make o2cb fail to manage nodes in low level, and then cause ocfs2 
> > not work. 
> > 
> > In my end,I use corosync/pacemaker  cluster stack. Nodes in cluster must  
> ssh 
> > each other without passwd. 
> > Hope it can help. 
> > 
> > Also,please add the mail list ;-). 
> > 
> > -- 
> > Best regards, 
> > Zhen, Ren 
> > HA team, SUSE 
> > 
> > 
> >>>> 
> >> There's one problem though. 
> >> 
> >> I tried on another node this method, but result is different. Its 
> >> network interface (em2) does not seem to be able to reach first node. 
> >> 
> >> Mar 17 12:54:23 app2 kernel: [   21.952527] igb: em2 NIC Link is Up 
> >> 1000 Mbps Full Duplex, Flow Control: RX/TX 
> >> Mar 17 12:54:23 app2 kernel: [   21.952737] IPv6: 
> >> ADDRCONF(NETDEV_CHANGE): em2: link becomes ready 
> >> Mar 17 12:54:28 app2 kernel: [   26.756863] OCFS2 Node Manager 1.5.0 
> >> Mar 17 12:54:28 app2 kernel: [   26.760481] OCFS2 DLM 1.5.0 
> >> Mar 17 12:54:28 app2 kernel: [   26.761550] ocfs2: Registered cluster 
> >> interface o2cb 
> >> Mar 17 12:54:28 app2 kernel: [   26.770524] OCFS2 DLMFS 1.5.0 
> >> Mar 17 12:54:28 app2 kernel: [   26.770625] OCFS2 User DLM kernel 
> >> interface loaded 
> >> Mar 17 12:54:35 app2 kernel: [   33.819716] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:36 app2 kernel: [   34.882978] OCFS2 1.5.0 
> >> Mar 17 12:54:38 app2 kernel: [   36.822648] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:41 app2 kernel: [   39.825565] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:44 app2 kernel: [   42.828493] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:47 app2 kernel: [   45.831420] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:50 app2 kernel: [   48.834359] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:53 app2 kernel: [   51.837318] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:56 app2 kernel: [   54.840224] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:54:59 app2 kernel: [   57.843135] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:55:01 app2 ntpdate[2490]: Can't find host ntp.ubuntu.com: 
> >> Name or service not known (-2) 
> >> Mar 17 12:55:01 app2 ntpdate[2490]: no servers can be used, exiting 
> >> Mar 17 12:55:02 app2 kernel: [   60.846069] o2net: Connection to node 
> >> web1 (num 1) at 192.168.56.100:7777 shutdown, state 7 
> >> Mar 17 12:55:02 app2 kernel: [   60.850050] o2net: No connection 
> >> established with node 1 after 30.0 seconds, giving up. 
> >> Mar 17 12:55:20 app2 ntpdate[3496]: no server suitable for synchronization 
> >> found 
> >> Mar 17 12:55:34 app2 kernel: [   92.881306] o2net: No connection 
> >> established with node 1 after 30.0 seconds, giving up. 
> >> Mar 17 12:55:36 app2 kernel: [   95.179544] o2cb: This node could not 
> >> connect to nodes: 1. 
> >> Mar 17 12:55:36 app2 kernel: [   95.184416] o2cb: Cluster check 
> >> failed. Fix errors before retrying. 
> >> Mar 17 12:55:36 app2 kernel: [   95.189351] 
> >> (mount.ocfs2,3465,3):ocfs2_dlm_init:3004 ERROR: status = -107 
> >> Mar 17 12:55:36 app2 kernel: [   95.194479] 
> >> (mount.ocfs2,3465,3):ocfs2_mount_volume:1881 ERROR: status = -107 
> >> Mar 17 12:55:36 app2 kernel: [   95.199577] ocfs2: Unmounting device 
> >> (252,2) on (node 0) 
> >> Mar 17 12:55:36 app2 kernel: [   95.199589] 
> >> (mount.ocfs2,3465,3):ocfs2_fill_super:1229 ERROR: status = -107 
> >> 
> >> 2015-03-17 12:29 GMT+08:00 Zhen Ren <zren at suse.com>: 
> >> > Congratulations! 
> >> > I also learned something from you. Thanks for your share. 
> >> > 
> >> > 
> >> > -- 
> >> > Best regards, 
> >> > Zhen, Ren 
> >> > HA team, SUSE 
> >> 
> >> 
> > 
>  
>