Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase

Question on installing new active node in an active-passive cluster configuration after current active node has failed Expand / Collapse
Author
Message
Posted Wednesday, February 26, 2014 6:15 PM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Yesterday @ 8:53 AM
Points: 1,861, Visits: 3,599
We have an active-passive cluster with 2 nodes, A and B, hosting a failover-cluster SQL instance (FCI).

When first installing the FCI, we first installed it on A and then used the "Add node to cluster" feature to install it on B.

So A was the "active" and B the "passive" node in the cluster configuration of the FCI.

Now A has been corrupted and we need to rebuild the server and re-install the SQL on A.

Should I be using the "Add node to cluster" feature to re-install the SQL on the rebuilt A server?

So my question is:
how do I re-install the SQL failover-cluster instance on a new node that is substituting a failed "ACTIVE" node in a cluster?


__________________________________________________________________________________

Turbocharge Your Database Maintenance With Service Broker: Part 2
Turbocharge Your Database Maintenance With Service Broker: Part 1
Real-Time Tracking of Tempdb Utilization Through Reporting Services
Monitoring Database Blocking Through SCOM 2007 Custom Rules and Alerts
Preparing for the Unthinkable - a Disaster/Recovery Implementation
Post #1545656
Posted Wednesday, February 26, 2014 6:36 PM
Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Thursday, July 24, 2014 11:26 AM
Points: 381, Visits: 1,857
When you said A has been corrupted, you mean databases?
Post #1545659
Posted Thursday, February 27, 2014 3:32 AM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Yesterday @ 8:53 AM
Points: 1,861, Visits: 3,599
No, the server itself is unusable; something happened and we are no longer able to even remote-desktop to it.

It needs to be rebuilt and the SQL re-installed. Problem is this was initially the "active" node in the cluster configuration.

The SQL instance now resides in the "passive" node, B, but we have no HA for it; if something happens to server B, we have no failover option at the moment.


__________________________________________________________________________________

Turbocharge Your Database Maintenance With Service Broker: Part 2
Turbocharge Your Database Maintenance With Service Broker: Part 1
Real-Time Tracking of Tempdb Utilization Through Reporting Services
Monitoring Database Blocking Through SCOM 2007 Custom Rules and Alerts
Preparing for the Unthinkable - a Disaster/Recovery Implementation
Post #1545777
Posted Thursday, February 27, 2014 3:49 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 1:50 PM
Points: 6,191, Visits: 13,340
Marios Philippopoulos (2/26/2014)
We have an active-passive cluster with 2 nodes, A and B, hosting a failover-cluster SQL instance (FCI).

When first installing the FCI, we first installed it on A and then used the "Add node to cluster" feature to install it on B.

So A was the "active" and B the "passive" node in the cluster configuration of the FCI.

Now A has been corrupted and we need to rebuild the server and re-install the SQL on A.

Should I be using the "Add node to cluster" feature to re-install the SQL on the rebuilt A server?

So my question is:
how do I re-install the SQL failover-cluster instance on a new node that is substituting a failed "ACTIVE" node in a cluster?

Did you evict the node from the cluster or remove it cleanly or did you just shut it down and rebuild it?


-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1545783
Posted Thursday, February 27, 2014 7:43 AM
Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Thursday, July 24, 2014 11:26 AM
Points: 381, Visits: 1,857
Marios Philippopoulos (2/27/2014)
No, the server itself is unusable; something happened and we are no longer able to even remote-desktop to it.

It needs to be rebuilt and the SQL re-installed. Problem is this was initially the "active" node in the cluster configuration.

The SQL instance now resides in the "passive" node, B, but we have no HA for it; if something happens to server B, we have no failover option at the moment.


So you have all the databases live on the other node which was passive before active server went down. So failover happened to passive node before the active went down, right? If yes, the cluster can be on the single node. Try to work on the other node and fix it. Try to run the remove node from sql installation and then evict the node and start from scratch.
Post #1545892
Posted Thursday, February 27, 2014 8:09 AM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Yesterday @ 8:53 AM
Points: 1,861, Visits: 3,599
Perry Whittle (2/27/2014)
Marios Philippopoulos (2/26/2014)
We have an active-passive cluster with 2 nodes, A and B, hosting a failover-cluster SQL instance (FCI).

When first installing the FCI, we first installed it on A and then used the "Add node to cluster" feature to install it on B.

So A was the "active" and B the "passive" node in the cluster configuration of the FCI.

Now A has been corrupted and we need to rebuild the server and re-install the SQL on A.

Should I be using the "Add node to cluster" feature to re-install the SQL on the rebuilt A server?

So my question is:
how do I re-install the SQL failover-cluster instance on a new node that is substituting a failed "ACTIVE" node in a cluster?

Did you evict the node from the cluster or remove it cleanly or did you just shut it down and rebuild it?


We have not done anything yet to fix this.
We have the SQL instance running fine on the passive node B; it had failed over cleanly when node A failed.
Assuming we can get a new server added to the cluster, how should I re-install the SQL on that node?

Use the "Add another node" installation option? Would that work?


__________________________________________________________________________________

Turbocharge Your Database Maintenance With Service Broker: Part 2
Turbocharge Your Database Maintenance With Service Broker: Part 1
Real-Time Tracking of Tempdb Utilization Through Reporting Services
Monitoring Database Blocking Through SCOM 2007 Custom Rules and Alerts
Preparing for the Unthinkable - a Disaster/Recovery Implementation
Post #1545909
Posted Thursday, February 27, 2014 8:14 AM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Yesterday @ 8:53 AM
Points: 1,861, Visits: 3,599
muthyala_51 (2/27/2014)
Marios Philippopoulos (2/27/2014)
No, the server itself is unusable; something happened and we are no longer able to even remote-desktop to it.

It needs to be rebuilt and the SQL re-installed. Problem is this was initially the "active" node in the cluster configuration.

The SQL instance now resides in the "passive" node, B, but we have no HA for it; if something happens to server B, we have no failover option at the moment.


So you have all the databases live on the other node which was passive before active server went down. So failover happened to passive node before the active went down, right? If yes, the cluster can be on the single node. Try to work on the other node and fix it. Try to run the remove node from sql installation and then evict the node and start from scratch.


Correct.

When you say "start from scratch", do you mean add a fresh new node to the cluster and use the "Add another node" installation option on the new node to install another "passive" node for the FCI?

I am just not sure about the specifics of what to do once a new node has been added to the cluster (to replace the failed one, which we cannot fix, BTW) and I am ready to install the failover-cluster instance on that new node. Bear in mind, I already have the FCI running on node B.


__________________________________________________________________________________

Turbocharge Your Database Maintenance With Service Broker: Part 2
Turbocharge Your Database Maintenance With Service Broker: Part 1
Real-Time Tracking of Tempdb Utilization Through Reporting Services
Monitoring Database Blocking Through SCOM 2007 Custom Rules and Alerts
Preparing for the Unthinkable - a Disaster/Recovery Implementation
Post #1545911
Posted Thursday, February 27, 2014 9:04 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 1:50 PM
Points: 6,191, Visits: 13,340
Marios Philippopoulos (2/27/2014)
Perry Whittle (2/27/2014)
Marios Philippopoulos (2/26/2014)
We have an active-passive cluster with 2 nodes, A and B, hosting a failover-cluster SQL instance (FCI).

When first installing the FCI, we first installed it on A and then used the "Add node to cluster" feature to install it on B.

So A was the "active" and B the "passive" node in the cluster configuration of the FCI.

Now A has been corrupted and we need to rebuild the server and re-install the SQL on A.

Should I be using the "Add node to cluster" feature to re-install the SQL on the rebuilt A server?

So my question is:
how do I re-install the SQL failover-cluster instance on a new node that is substituting a failed "ACTIVE" node in a cluster?

Did you evict the node from the cluster or remove it cleanly or did you just shut it down and rebuild it?


We have not done anything yet to fix this.
We have the SQL instance running fine on the passive node B; it had failed over cleanly when node A failed.
Assuming we can get a new server added to the cluster, how should I re-install the SQL on that node?

Use the "Add another node" installation option? Would that work?

OK, evict the node from the cluster and perform a clean rebuild. When you reinstall you use the add node wizard to join the existing instance running on node B


-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1545950
Posted Thursday, February 27, 2014 9:10 AM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Yesterday @ 8:53 AM
Points: 1,861, Visits: 3,599
Thank you!

__________________________________________________________________________________

Turbocharge Your Database Maintenance With Service Broker: Part 2
Turbocharge Your Database Maintenance With Service Broker: Part 1
Real-Time Tracking of Tempdb Utilization Through Reporting Services
Monitoring Database Blocking Through SCOM 2007 Custom Rules and Alerts
Preparing for the Unthinkable - a Disaster/Recovery Implementation
Post #1545955
« Prev Topic | Next Topic »

Add to briefcase

Permissions Expand / Collapse