Albert Widjaja
asked on
VMotion of Exchange mailbox server causing downtime
Hi,
When doing VMotion for Exchange Server 2007 SP1 CCR mailbox in the DR instance, it stopped the production node and causing downtime to the production mailbox server eventhough the VMotion is only for the DR instance ?
Because during the VMotion the Virtual IP and clustername is not responding to ping, thus causing massive total outage in the production mailbox.
The solution was to turn off the DR instance after the VMotion to make it fail over to the production node again.
When doing VMotion for Exchange Server 2007 SP1 CCR mailbox in the DR instance, it stopped the production node and causing downtime to the production mailbox server eventhough the VMotion is only for the DR instance ?
Because during the VMotion the Virtual IP and clustername is not responding to ping, thus causing massive total outage in the production mailbox.
The solution was to turn off the DR instance after the VMotion to make it fail over to the production node again.
I would say it depends on how your storage is setup. Vmotion should only (maybe) drop the ping for 1-2 seconds but when you are moving it to the other server do you have a seperate iSCSI LUN attatched or is everything stored on a central location (SAN).
ASKER
It is connected to the same Fiber channel SAN as and te network as well.
It's possible this can happen, that you may lose a "few pings" whilst the vMotion is in progress, but this depends on how busy your Exchange Server.
wht vsphere version are you using ?
1. how you attached the hdd on exchange server..because i face this issue as i add as d rive the esx local hdd instead of storage LUN.
2.http://www.vmware.com/files/pdf/resources/availability_recovery_exchange_sb.pdf
3.http://communities.vmware.com/message/1536600
1. how you attached the hdd on exchange server..because i face this issue as i add as d rive the esx local hdd instead of storage LUN.
2.http://www.vmware.com/files/pdf/resources/availability_recovery_exchange_sb.pdf
3.http://communities.vmware.com/message/1536600
ASKER
I'm using VMware vSphere ESX 4.1.0 433742
the drives area all RDM presented from my EMC CX3 array, all of the ESX servers got access to the LUNs.
the drives area all RDM presented from my EMC CX3 array, all of the ESX servers got access to the LUNs.
ASKER
So the problem was after the VMotion of te Exchange Mailbox server in DR finished, all of my users got disconnected from the Exhcange Server.
Eventhough
Eventhough
ASKER
Eventhough the DR mailbox server took over but none of my users are able to connect back to the Outlook.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
I concur with Sembee2, we do not use vMotion on busy Exchange servers, that are heavily loaded, the STUN cycle can cause issues.
ASKER
But I only vMotion the passive node ?
I'm confused as to why it affects the production node.
I'm confused as to why it affects the production node.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.