Re: Detecting Record Lock Conditions

["Buck" <buck.calabro@xxxxxxxxxxxx>]

> I'm trying to have a program detect if a record is locked.
> If the program does find that the record it's trying to read
> is locked, it will wait 10 seconds before retrying to read
> the same record from the file.

The code you posted looks like it will do what you want.  Beware of a deadly
embrace though.  Job A locks record 1 in file A.  Job B locks record 1 in
file B, then tries to read file A.  Record is locked, goes into retry loop.
Job A tries to read file B.  Record is locked, goes into retry loop.  Each
job waits for the other job to release it's lock.

An option to consider is to do an automatic retry for n number of attempts,
or n minutes before 'giving up.'  Giving up can mean your program posting a
message (not the system error handler) to QSYSOPR to have the operator clear
the lock and then continue.  Your code can read the response and begin
another round of retry loop.

You can guess how I came to use this particular construct.  Timing is
EVERYthing.
  --buck