[Pacemaker] cluster doesn't failover - log at the end of msg

Thomas Schneider thomas.schneider at euskill.com
Tue Nov 3 17:28:29 EST 2009


Hi

Thanks for your help i solve the problem by replacing this line

location cli-standby-nfs nfs \
        rule $id="cli-standby-rule-nfs" -inf: #uname eq
storage02.myriapulse.local

by

location ms_drbd_nfs-master-storage01.myriapulse.local ms_drbd_nfs rule
role=master 100: \#uname storage01.myriapulse.local

but i have an other problem because i have 2 network card

                 Eth1: replication link
 --------------      ----------------
 | Server 1   |------| Server 2     |
 --------------      ----------------
      |                    |
     Eth0                 eth0
      |                    |  
         External Network

The data of the nfs server is accessed via eth0, but if y shutdown eth0 on
the node witch is running the resource, the resource doesn't migrate and is
unreachable, i think it's because of the replication link (who stay up)

node storage01.myriapulse.local \
        attributes standby="off"
node storage02.myriapulse.local \
        attributes standby="off"
primitive drbd_nfs ocf:linbit:drbd \
        params drbd_resource="nfs" \
        op monitor interval="15s" \
        meta target-role="Started"
primitive fs_nfs ocf:heartbeat:Filesystem \
        params device="/dev/drbd/by-res/nfs" directory="/share"
fstype="ext3" \
        meta is-managed="true"
primitive ftp-server lsb:proftpd \
        op monitor interval="1min"
primitive ip_nfs ocf:heartbeat:IPaddr2 \
        params ip="10.1.1.69" nic="eth0"
primitive nfs-kernel-server lsb:nfs-kernel-server \
        op monitor interval="1min"
group nfs fs_nfs ip_nfs nfs-kernel-server \
        meta target-role="Started"
ms ms_drbd_nfs drbd_nfs \
        meta master-max="1" master-node-max="1" clone-max="2"
clone-node-max="1" notify="true"
location drbd-fence-by-handler-ms_drbd_nfs ms_drbd_nfs \
        rule $id="drbd-fence-by-handler-rule-ms_drbd_nfs" $role="Master"
-inf: #uname ne storage02.myriapulse.local
location ms_drbd_nfs-master-storage01.myriapulse.local ms_drbd_nfs \
        rule $id="ms_drbd_nfs-master-storage01.myriapulse.local-rule"
$role="master" 100: #uname eq storage01.myriapulse.local
colocation ftp_on_nfs inf: ftp-server nfs
colocation nfs_on_drbd inf: nfs ms_drbd_nfs:Master
order ftp_after_nfs inf: nfs ftp-server
order nfs_after_drbd inf: ms_drbd_nfs:promote nfs:start
property $id="cib-bootstrap-options" \
        dc-version="1.0.5-unknown" \
        cluster-infrastructure="openais" \
        expected-quorum-votes="2" \
        no-quorum-policy="ignore" \
        stonith-enabled="false"

Thanks for your help
Thomas Schneider


-----Message d'origine-----
De : drbd-user-bounces at lists.linbit.com
[mailto:drbd-user-bounces at lists.linbit.com] De la part de Lars Ellenberg
Envoyé : samedi 31 octobre 2009 10:15
À : drbd-user at lists.linbit.com
Objet : Re: [DRBD-user] cluster doesn't failover - log at the end of msg

On Sat, Oct 31, 2009 at 02:57:37AM +0100, Thomas Schneider wrote:
> Hello,
> 
> I'm trying to setup a cluster for shared storage with pacemaker, drbd, and
> nfs.
> I use two server and there is two network interface one each server, eth0
> connected to the network, and eth1 direct link betwene the two node for
drbd
> replication,


> I can migrate the resource between the to node, but the problem is when i
> make a hard power-off of server01 ( the server where is running the
> resource) the second server (server02) doesn't make failover. (the
resource
> doesn't start) May you can take a look to my config file in the following


> storage01:~# crm configure show
> node storage01.myriapulse.local \
>         attributes standby="off"
> node storage02.myriapulse.local \
>         attributes standby="off"
> primitive drbd_nfs ocf:linbit:drbd \
>         params drbd_resource="nfs" \
>         op monitor interval="15s" \
>         meta target-role="Started"
> primitive fs_nfs ocf:heartbeat:Filesystem \
>         params device="/dev/drbd/by-res/nfs" directory="/share"
> fstype="ext3" \
>         meta is-managed="true"
> primitive ftp-server lsb:proftpd \
>         op monitor interval="1min"
> primitive ip_nfs ocf:heartbeat:IPaddr2 \
>         params ip="10.1.1.69" nic="eth0"
> primitive nfs-kernel-server lsb:nfs-kernel-server \
>         op monitor interval="1min"
> group nfs fs_nfs ip_nfs nfs-kernel-server \
>         meta target-role="Started"
> ms ms_drbd_nfs drbd_nfs \
>         meta master-max="1" master-node-max="1" clone-max="2"
> clone-node-max="1" notify="true"
> location cli-standby-nfs nfs \
>         rule $id="cli-standby-rule-nfs" -inf: #uname eq
> storage02.myriapulse.local colocation ftp_on_nfs inf: ftp-server nfs
> colocation nfs_on_drbd inf: nfs ms_drbd_nfs:Master order ftp_after_nfs
inf:
> nfs ftp-server order nfs_after_drbd inf: ms_drbd_nfs:promote nfs:start

You really should get your line breaks right!

> property $id="cib-bootstrap-options" \
>         dc-version="1.0.5-unknown" \
>         cluster-infrastructure="openais" \
>         expected-quorum-votes="2" \
>         no-quorum-policy="ignore" \
>         stonith-enabled="false"

> Oct 31 02:47:06 storage02 kernel: [419458.084631] block drbd0:
helpercommand: /sbin/drbdadm fence-peer minor-0
> Oct 31 02:47:07 storage02 crm-fence-peer.sh[24594]: invoked for nfs
> Oct 31 02:47:07 storage02 kernel: [419459.283186] block drbd0:
helpercommand: /sbin/drbdadm fence-peer minor-0 exit code 4 (0x400)
> Oct 31 02:47:07 storage02 kernel: [419459.283186] block drbd0:
fence-peerhelper returned 4 (peer was fenced)
> Oct 31 02:47:07 storage02 kernel: [419459.283186] block drbd0:
role(Secondary -> Primary ) pdsk( DUnknown -> Outdated ) 
> Oct 31 02:47:07 storage02 kernel: [419459.343187] block drbd0: Creating
newcurrent UUID
> Oct 31 02:47:07 storage02 lrmd: [23822]: info: RA
output:(drbd_nfs:0:promote:stdout) 
> Oct 31 02:47:08 storage02 crmd: [23825]: info: process_lrm_event:
LRMoperation drbd_nfs:0_promote_0 (call=12, rc=0,
cib-update=42,confirmed=true) complete ok

DRBD is promoted just fine.

The resources using it are not.
Which is expected:

As long as you explicitly forbid "nfs" to run on storage02,
you should not complain about nfs not being started on storage02.

Hint:

> location cli-standby-nfs nfs \
>         rule $id="cli-standby-rule-nfs" -inf: #uname eq
storage02.myriapulse.local


-- 
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and consulting http://www.linbit.com

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.
__
please don't Cc me, but send to list   --   I'm subscribed
_______________________________________________
drbd-user mailing list
drbd-user at lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user




More information about the Pacemaker mailing list