Cas x3 MBX x 4 Exchange 2013 CU7

Memory of one CAS wasn't working due to high load and then brought several conponents done.

RPC proxy and EWS, Active Sync was inactive, so users couldn't access by outlook and iPhone.

I rebooted the CAS server once, and above conponents was retuning to active.

And one hour later, SCOM alerts as below.

The Outlook.Proxy health set has detected a problemThe health manager is reporting that recycling the
MSExchangeRpcProxyFrontEndAppPool app pool has failed to restore health and
attempts to mark the protocol offline have failed

Anyone faced the same issue?

and soultion_?

• Edited by tanale Tuesday, June 30, 2015 4:51 AM

Sorry my late response.

Yes CPU is ok, but it's tendency that memory utilization become 80% to 90% and then stop the components.

Is there possibility if Exchange 2013 high availability management works out?

• Edited by tanale 49 minutes ago

Hi,

How are you load balancing the CAS servers.

Exchange HA only accounts for Mailbox Role.

                   

It's using Load balancer for three CAS servers.

It looks no issues for load balancing.

Sorry my late response.

Yes CPU is ok, but it's tendency that memory utilization become 80% to 90% and then stop the components.

Is there possibility if Exchange 2013 high availability management works out?

• Edited by tanale Monday, July 06, 2015 6:39 AM

Hi Tanale,

It could be Exchange workload management kicking in for low Resources, look for events in event logs as mentioned by Simon.

Rather than completely blocking a user from performing operations, throttling occurs and processes are delayed for short periods of time (think of them as "microdelays"), occurring before they cause a significant impact on a server.

There was a bug in one of the iPhone iOS versions causing memory leak for earlier Exchange versions, make sure unsuppported versions are disabled for active sync.