2618949 - Work processes unable to restart

1. A work process is seen without status at the transaction SM50.

Its PID changes at every few seconds, but the corresponding trace file (dev_wXX) is not being updated.

The counter at the "Err" column of the corresponding work process is increasing.

2. At operating system level (SAP MMC or dpmon), the status of the affected process might be "NEW".

3. The Dispatcher trace (dev_disp) logs that the affected process is dying:

<date and time>
*** ERROR => DpHdlDeadWp: WX (pid YYYY) died (severity=0, status=0) [dpxxwp.c     1748]

<date and time>
*** ERROR => DpHdlDeadWp: WX (pid ZZZZZ) died (severity=0, status=0) [dpxxwp.c     1748]

4. The "stderrX" file corresponding to the Dispatcher process has entries similar to the below (the values in red might vary):

NTSA_VirtualAlloc : inconsistent size for id 1  (expected 0000000100000000, got 000000017E600000)

The issue can affect one or more work processes at the same time. It could even reach a point where all work processes are affected and the instance becomes unusable.

• Product independent
• SAP Release independent
• Client/Server Technology - Memory Management
• SAP running on Windows

NTSA_VirtualAlloc, Inconsistent size, blank status, status empty , KBA , BC-CST-MM , Memory Management , BC-CST , Client/Server Technology , BC-OP-NT , Windows , Problem