Has anyone seen this Unusual NLB and vmware problem
We have a 2008 NLB cluster of 2 machines we shall call A and B and a 2008 file server called C
They are all on VMware as virtual machines. The cluster servers are sharepoint servers and access data from fileserver C.
The problem we had was that cluster server B could not see fileserver C and vice versa either to ping or tracert or to access shares via IP or host name.
but could they could both see all other servers on the network,
After trying lots of things we tried performing a drain stop on server A from NLB manager. Suddenly Cluster server B could see the file server and vice versa. We brought back cluster server A and still everything worked fine.
Thing is that while it works for now we need to know why this happened before we can make these servers live in a production environment but are at a loss to explain.
They are all on VMware as virtual machines. The cluster servers are sharepoint servers and access data from fileserver C.
The problem we had was that cluster server B could not see fileserver C and vice versa either to ping or tracert or to access shares via IP or host name.
but could they could both see all other servers on the network,
After trying lots of things we tried performing a drain stop on server A from NLB manager. Suddenly Cluster server B could see the file server and vice versa. We brought back cluster server A and still everything worked fine.
Thing is that while it works for now we need to know why this happened before we can make these servers live in a production environment but are at a loss to explain.
bgoering
Is the fileserver on the same network segment as the nlb servers? Are you using nlb multicast?
Also take a look at the hotfix info from http://support.microsoft.com/kb/960916
ASKER
The file server and cluster are on the same subnet and on the same esx host, they are using unicast.
We are thinking maybe using a virtual switch on vmware and using that to load balance maybe.
I am just looking at the hotfix right now, thanks.
We are thinking maybe using a virtual switch on vmware and using that to load balance maybe.
I am just looking at the hotfix right now, thanks.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
use vSwitch for load balance instead routing traffic to physical switch.
I had same type of issue where sharepoint server i was able to access remote but withing sharepoint environment it was not connecting each other ....event ping was replying with alive connection.
after reporting to network and security team they add that particular server Mac address and Server IP address in switch.
This sharepoint environment was in VLAN...hope this will help.
after reporting to network and security team they add that particular server Mac address and Server IP address in switch.
This sharepoint environment was in VLAN...hope this will help.