I currently have my BES 5 install as a VM on Hyper-V (only have about 30 users right now). I am wanting to cluster my Hyper-V for high availability. Will this affect BES in any form or fasion?

Nope. Works like a charm. Only in large environments you want to keep an eye on hardware resources.

Hi There,

I recently deployed a Hyper-V cluster and after deployement everything seemed to work fine except BAS. WIth Both nodes up only one would let you logon, the other (the last one to start) would prompt you to logon but not let you in. The BAS log was littered with events like this:

(05/26 14:17:32:575):{Timer-11} [com.rim.bes.bas.singletondeploymentbarrier.SingletonDeploymentBarrierManagerBean] [INFO] [BBAS-1011] {u=SystemUser, t=7732} isCurrentNodeTheSingletonInNewTransactionLocal CLUSTER: heartbeat from another server (SERVER01.MYDOMAIN.COM) is still active, this server (SERVER02.MYDOMAIN.COM) cannot take over as singleton------------------------------------------

By Default (anyone correct me if I've bunched the explanation I've been out of town for a week and the Brain is a little foggy) BES HA uses UDP for inter cluster communication (I think this is only for the JBOSS components, BAS, Web Desktop, MDS). While both my nodes were on the same subnet they were deployed on Highly available VMs on two different Hyper-V Hosts that are a part of a 6-way cluster deployed on CSV.

In my deployment it looks like the Hyper-V Virtual Switches were not passing UDP packets properly and once we switched to TCP using the RIM KB article below everything worked.

A few words of note, you have a lot of little edits to do backup the XML files before you change them and/or create a checkpoint of your VMs prior to updating just in case you mess something up.

Not sure if you'll see the same problem (please post if you do so I know I'm not the only one ) but if you do I hope this helps


KB19654 - View Document