Resolving a stoppage or loop
A stoppage is a situation where the server does not start or complete a function and is not using any microprocessor power.
A stoppage might be just one session or process that is not processing, or it can be the entire IBM Storage Protect server not responding. A loop is a situation where no progress is being made, but the server is using a high amount of microprocessor power. A loop can affect just one session or process, or it can affect the entire server.
You might collect documentation to resolve this type of problem, depending on whether the server
is able to respond to commands. A Perl
script is available for you to collect server data. It is helpful to schedule the
SHOW command list to run intermittently so that you can then see the behavior
that precedes the stoppage situation.
- For a stoppage or a loop where the server can respond to commands, issue the following commands
to help determine the cause of the stoppage:
- QUERY SESSION f=d
- QUERY PROCESS
- SHOW RESQ
- SHOW THREADS
- SHOW DEADLOCK
- SHOW TXNT
- SHOW DBTXNT
- SHOW LOCKS
- SHOW LIBR
- SHOW MP
- SHOW SESS
- SHOW ASQ
- SHOW ASVOL
- SHOW DBV
- SHOW SSS
- SHOW CSV (Issue this command only when the problem is related to scheduling.)
- When a server hangs or loops, issue the following commands to provide a detailed diagnostic
snapshot of the IBM
Storage Protect
environment:
You can use the db2support.zip file that is generated for troubleshooting.db2fodc -hang -alldbs db2support . -d database -s - In addition to the output from the listed commands, or in the cases of a server that cannot
respond to commands, collect a dump. The way that you collect a dump depends on the operating system.

Issue the KILL -11 command on the dsmserv process to
create a core file. To run the kill
command, obtain the process ID by issuing the PS command.
Search for collecting user mode dumps at the Microsoft website at http://support.microsoft.com/.