加载中…惠普刀片机7代及虚拟VSpare5（十一）Cluster
(2012-09-27 23:49:32)
标签：
it |
分类： 工作 |
I setup two 2 node clusters, one of them went
smoothly and another one automatically come with a quorum
disk. We do not need the Quorum disk duo to the
disks limitation. I followed the steps to change quorum to "node
mojority"to free up the disk.
1. Open Failover Cluster Manager + connect to cluster
2. Configure cluster quorum settings
3. Select Node Majority
it says move back the disk to available storage. but it is not. so the disk 1 is stocked on the cluster group. I did most everything to bring it to available storage. but failed.
So, delete the cluster and try to recreate it. bad thing is I did not do it from GUI and run into command directly.
Use the cluster.exe command to delete a Server 2008 failover cluster, and the /cleanup switch to delete any Active Directory (AD) information. In the following example, I obtained a list of existing clusters and deleted one of the listed clusters. Then, I reopened the list to confirm that the deletion was successful.
To obtain a list of existing clusters, I used the command
cluster /list
The output was:
Cluster Name
---------------
CLUS01
CLUS02
Then, I destroyed the CLUS02 cluster by using the command
cluster CLUS02 /destroy /cleanup
The output was
Are you sure you want to destroy this cluster? (Y/N):
I typed Y and pressed Enter. The output was
Destroying cluster CLUS02...
11% Moving resource group 'Cluster Group'.
22% Deleting resource group 'Available Storage'.
33% Cleaning up computer objects for cluster 'CLUS02'.
44% Evicting node Node2 from cluster.
55% Attempting to restore node Node2 to its non-clustered state.
66% Taking resource group 'Cluster Group' offline.
77% Cleaning up resources in core Cluster groups.
88% Evicting node Node1 from cluster.
100% Attempting to restore node Node1 to its non-clustered state.
To reopen the cluster list, I used the command
cluster /list
The output was
Cluster Name
---------------
CLUS01
A cluster should be in a state in which it can be started to clean up AD information. If your cluster nodes are in a state in which the cluster you want to delete can't be started, use the /forcecleanup switch to force a cluster-information cleanup, as in the following sample code:
"cluster node (node) /forcecleanup"
It should be end of the story. But, after I recreate the cluster, the service not going to start and keep complaining. Then I decide to remove the feature and reboot the OS and try again.
When I try remove the Failover Cluster feature has following message:
Cannot remove Failover Clusting
This server is an active node in a fail-over cluster. Uninstalling the Failover CVlustering feature on those node may impact the availability of clustered service and applications. It is recommended that you first evict the server from cluster membership. This can be done through the Failover Cluster Management snap-in by expanding the console tree under Nodes, selecting the node, clicking More Actions, and then clicking Evict.
Now, we need new tools called powershell to help me out.
Run powershell as an administrator
Import-Module FailoverClusters
then clear-clusternode
After this, I managed to remove cluster features from both nodes and re add it. one of node is working as expected, another one gave me an error saying "computer is joined to a cluster". Then I did the above cycle for that node and still have same issue. I guess that maybe got something to do with AD. I run out of the time and change the host name and re-do it. it works now.
1. Open Failover Cluster Manager + connect to cluster
2. Configure cluster quorum settings
3. Select Node Majority
it says move back the disk to available storage. but it is not. so the disk 1 is stocked on the cluster group. I did most everything to bring it to available storage. but failed.
So, delete the cluster and try to recreate it. bad thing is I did not do it from GUI and run into command directly.
Use the cluster.exe command to delete a Server 2008 failover cluster, and the /cleanup switch to delete any Active Directory (AD) information. In the following example, I obtained a list of existing clusters and deleted one of the listed clusters. Then, I reopened the list to confirm that the deletion was successful.
To obtain a list of existing clusters, I used the command
cluster /list
The output was:
Cluster Name
---------------
CLUS01
CLUS02
Then, I destroyed the CLUS02 cluster by using the command
cluster CLUS02 /destroy /cleanup
The output was
Are you sure you want to destroy this cluster? (Y/N):
I typed Y and pressed Enter. The output was
Destroying cluster CLUS02...
11% Moving resource group 'Cluster Group'.
22% Deleting resource group 'Available Storage'.
33% Cleaning up computer objects for cluster 'CLUS02'.
44% Evicting node Node2 from cluster.
55% Attempting to restore node Node2 to its non-clustered state.
66% Taking resource group 'Cluster Group' offline.
77% Cleaning up resources in core Cluster groups.
88% Evicting node Node1 from cluster.
100% Attempting to restore node Node1 to its non-clustered state.
To reopen the cluster list, I used the command
cluster /list
The output was
Cluster Name
---------------
CLUS01
A cluster should be in a state in which it can be started to clean up AD information. If your cluster nodes are in a state in which the cluster you want to delete can't be started, use the /forcecleanup switch to force a cluster-information cleanup, as in the following sample code:
"cluster node (node) /forcecleanup"
It should be end of the story. But, after I recreate the cluster, the service not going to start and keep complaining. Then I decide to remove the feature and reboot the OS and try again.
When I try remove the Failover Cluster feature has following message:
Cannot remove Failover Clusting
This server is an active node in a fail-over cluster. Uninstalling the Failover CVlustering feature on those node may impact the availability of clustered service and applications. It is recommended that you first evict the server from cluster membership. This can be done through the Failover Cluster Management snap-in by expanding the console tree under Nodes, selecting the node, clicking More Actions, and then clicking Evict.
Now, we need new tools called powershell to help me out.
Run powershell as an administrator
Import-Module FailoverClusters
then clear-clusternode
After this, I managed to remove cluster features from both nodes and re add it. one of node is working as expected, another one gave me an error saying "computer is joined to a cluster". Then I did the above cycle for that node and still have same issue. I guess that maybe got something to do with AD. I run out of the time and change the host name and re-do it. it works now.
喜欢
0
赠金笔
后一篇：关于kumon