cluster 其中一台已經停機,如何讓另一台可以正常運作

[watermelonyu]

我將rac2 停機

本來rac1 可以正常running,但是再重新開機後

我的cluster 起不來了,請問怎解???

on node 1 (rac1)

# cmruncl

cmruncl : Validating network configuration...

Not probing node rac2 as it is currently unreachable.

cmruncl : Network validation complete

Cannot reach node "rac2".

# cmviewcl

cmviewcl : Cannot talk to all the nodes.

Cluster does not appear to be up



CLUSTER STATUS

cluster1 unknown



NODE STATUS STATE

rac1 down unknown

rac2 unknown unknown

[watermelonyu]

你可以使用cmnoderun -n rac1 来试试

[watermelonyu]

# cmrunnode rac1

cmrunnode : Unable to communicate with a running cluster or with all nodes in the cluster.

cmrunnode : In order to use cmrunnode, the cluster must already be running on a subset of reachable nodes or else all c

luster nodes must be reachable.

cmrunnode : Issuing cmrunnode again may succeed.

[watermelonyu]

# cmviewcl

cmviewcl : Cannot talk to all the nodes.

Cluster does not appear to be up



CLUSTER STATUS

cluster1 unknown



NODE STATUS STATE

rac1 down unknown

rac2 unknown unknown

[watermelonyu]

在你的2 node cluster中, 是否以rac2為quorum server,

如果答案是肯定的, 那麼rac1要以cmruncl將整個cluster最上層的服務帶起時, 因vote數不夠, 自然無法將cluster service帶起.

[watermelonyu]

我有設定qourum disk on share disk

FIRST_CLUSTER_LOCK_PV /dev/dsk/c4t0d1

還有我要怎樣看cluster 的vote 數

[watermelonyu]

use "cmruncl -f -n node1"

to force cluster startup. then,

"cmrunnode -v node1"

then

"cmrunpkg -n node1 -v pkg_name"



to bring up the services

[watermelonyu]

印象中, cluster vote數沒辦法透過command來看.

[watermelonyu]

另外，请检查你系统里面的/etc/cmcluster/下面的*.ascii文件，我觉得这也可能有问题