STATE-RESET and STATE-AWARE AppServer agent becomes orphan with errors 8121 and 8077

STATE-RESET and STATE-AWARE AppServer Agent becomes orphan with errors 8121 and 8077 shown in the Broker log

STATE-RESET and STATE-AWARE AppServer connectingTimeout does not work as expected

AppServer Agent becomes orphaned when connectingTimeout value is less that time taken by srvrConnectProc to execute

STATE-RESET AppServer Agent becomes orphan when connectingTimeout value is less that time taken by srvrConnectProc to execute
STATE-RESET/STATE-AWARE modes when the connectingTimeout value is less than time taken by srvrConnectProc to execute, the Agent connected to client gets unmanaged.

AppServer Broker query does not show the Agent as it becomes an orphan process
asbman -q, does not list the agent anymore
The AppServer Agent (_proapsv.exe /_proapsv) process remains in the system memory
connectingTimeout is only valid for STATE-RESET and STATE-AWARE modes

When connectingTimeout is lesser than time taken by srvrConnectProc to finish, the Server thread in Broker:

• Is in CONNECTING mode, 
• Waits in getServerIPC() event for the 'connectingTimeout' period of time  
• Then raises an EVENT_TIMEOUT
• This causes the Server thread to transition to AVAILABLE mode. 

When srvrConnectProc finishes execution:

• The AVAILABLE Server thread receives a CONNECT message. 
• AVAILABLE mode transitions to OBLIVION as per design (a CONNECT event is not accepted in AVAILABLE state). 
• This causes the thread to be removed from serverPool, making it unmanageable.

Upgrade to OpenEdge 11.0 or later where the server thread waits in CONNECTING mode ignoring connectingTimeout until the Agent sends some message to the Broker.

For STATE-RESET and STATE-AWARE modes, either ensure that connectingTimeout is longer than the srvrConnectProc proceedure time, or don't use connectingTimeout.

Progress Article:

 What is the purpose of the killAgentAfterTimeout property