Users getting the following random error when connecting to manual pool. RDP is the only allowed protocol and MS Remote Desktop Client has no issue. To fix, I restart view agent service, logout, or connect using MS RDP client then disconenct.
Unable to launch from Pool <poolid> for user <domain>\<username> :
All responding machines are currently in use
---------------------------------------------------------------------------
Agent log shows the following:
2012-12-21T23:30:30.507-07:00 DEBUG (04E4-1594) <MessageFrameWorkDispatch> [wsnm_desktop] commandhandler::startSession: Received windows credentials for SSO {SESSION:f2e6-***-d4d0}
2012-12-21T23:30:30.507-07:00 DEBUG (04E4-1594) <MessageFrameWorkDispatch> [wsnm_desktop] startSession requested with too many unmatched sessions: connected=1, pending=0, portal=0, disconnected=0 {SESSION:f2e6-***-d4d0}
2012-12-21T23:30:30.507-07:00 DEBUG (04E4-1594) <MessageFrameWorkDispatch> [wsnm_desktop] startSession refused on MaxSession limit, user <my.domain.com>\<username> {SESSION:f2e6-***-d4d0}
------------------
Connection logs show:
2012-12-21T23:21:34.448-07:00 DEBUG (09E4-05E8) <TP-Processor2> [DesktopSessionImp] (SESSION:3a50-***-d2be) startSession failed, maximum sessions reached.
2012-12-21T23:21:34.449-07:00 DEBUG (09E4-05E8) <TP-Processor2> [EventLogger] (SESSION:3a50-***-d2be) Error_Event:[BROKER_POOL_OVERLOADED] "Unable to launch from Pool <mypoolid> for user <my.domain.com>\<username>: All responding machines are currently in use": Source=com.vmware.vdi.sessionclientapi.FarmImp, DesktopId=<mypoolid>, Time=Fri Dec 21 23:21:34 MST 2012, Severity=AUDIT_FAIL, Node=<myconnectionserver_fqdn>, PoolId=<mypoolid>, UserSID=S-1-5-21-1935655697-1682526488-725345543-1624, Module=Broker, UserDisplayName=<my.domain.com>\<username>, Acknowledged=true
2012-12-21T23:21:34.449-07:00 DEBUG (07A0-0CD8) <MessageFrameWorkDispatch> [MessageFrameWork] System::WriteWindowsEvent
2012-12-21T23:21:34.449-07:00 INFO (09E4-05E8) <TP-Processor2> [SessionLaunchContext] (SESSION:3a50-***-d2be) <my.domain.com>\<username>, App = <mypoolid>: Session request failed.
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (15ms): User has 1 possible sessions to analyze.
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (15ms): Possible session match found on server cn=68845310-87a2-4691-bec5-3cfe211b116f,ou=servers,dc=vdi,dc=vmware,dc=int.
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (15ms): No matching capids, selected session on server 68845310-87a2-4691-bec5-3cfe211b116f
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (15ms): Session is connected. Will disconnect first.
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (112ms): 1 server(s) were rejected because StartSession failure. Possible session clash. [68845310-87a2-4691-bec5-3cfe211b116f]
(SESSION:3a50-***-d2be:) [<my.domain.com>\<username>, App = PC] (113ms): Application launch failed, exception was: All available desktop sources for this desktop are currently busy. Please try again later.
----------------------------------
Is there some sort of timeout of something I can use to prevent this? It looks like Connection Servers missed a disconnect message somewhere and think a session is still going. In fact, the audit logs show this. Any ideas would be helpful. I am running 5.1 and the physical machines are not set to sleep. I have also tried excluding the agent folder from AV with no success. I have found no way to replicate this and it happens randomly.