cancel
Showing results for 
Search instead for 
Did you mean: 

shutdown and restart Cluster with VEA

mghomer
Level 6

Hi all,

I have to shutdown our Hardware for a while. Now I'm looking for the best way to do that.

I have two storage SAN's connected to two hardware micosoft clusternodes(Windows Server 2003) with VEA 3.2.

I will shutdown the sytsems this way:

1. passive Cluster node

2. aktiv cluster node

3. first san controller

4. second san controller

5. both storage's

6. FB switche

and the restart this way:

1. FB switche

2. both storage's

3. both storage san controller

4. both cluster node's

Are there any things i have to regard? Do the VEA start the to resync after the restart for hours? or it will only reconnect.

I hope you can help me.

Thank you!

3 REPLIES 3

Wally_Heim
Level 6
Employee

Hi mghomer,

I'm assuming that you have a mirror setup between the two storage SANs.  In general, SFW should automatically recover the mirror if needed when this configuration is restarted.

You did not mention why you need to shutdown everything, as long as you are not physically changing the luns seen by the cluster and used in the mirror then VEA will recover as expected.  However, if you are replacing the disk(s) that make up the luns from the array then you might need to do some manual items to resync the plex in that array.

Thank you,

Wally

mikebounds
Level 6
Partner Accredited

As you are shutting down the cluster then this will stop the application and therefore nothing will write to the storage while you shutdown and restart SAN.  Therefore there will be nothing to resync and therefore the cluster and storage will start as if nothing had happened.

VEA never actually does anything, this is just a GUI that connects to SFW running on a host - it is SFW that does the work.  So whether you are connected to VEA or not has no effect on resyncs etc.

If you have a mirrored Volume across 2 arrays and you loose an an array while application is writing then SFW will continue to write to array that is up and then when downed array returns, SFW will kick of resync automatically.  If you loose a LUN in the array (or downed array needs to be replaced), then like Wally says you will need to do some manual items to resync the plex in that array.

Mike


@mikebounds wrote:

As you are shutting down the cluster then this will stop the application and therefore nothing will write to the storage while you shutdown and restart SAN.  Therefore there will be nothing to resync and therefore the cluster and storage will start as if nothing had happened.

VEA never actually does anything, this is just a GUI that connects to SFW running on a host - it is SFW that does the work.  So whether you are connected to VEA or not has no effect on resyncs etc.track usps package

If you have a mirrored Volume across 2 arrays and you loose an an array while application is writing then SFW will continue to write to array that is up and then when downed array returns, SFW will kick of resync automatically.  If you loose a LUN in the array (or downed array needs to be replaced), then like Wally says you will need to do some manual items to resync the plex in that array. usps tracking

Mike


agreed wirh mike.