On Thu, 19 May 2011 08:56:49 -0500, Tim Deller wrote: >I would try coding a wait (for a few seconds at least) after the dataset close. >This would allow time for the deallocate and catalog update to complete. > Grrr. I detest designs that require the programmer to use delays to (hopefully) resolve race conditions.
The requirement that a program not proceed until data have been hardened is sufficiently common that there ought to be a "WAIT=YES" operand for the CLOSE macro such that it would not return until all reasonable side effects are complete. "Reasonable"? Waiting for a catalog update is reasonable; waiting until a tape has been transported to a remote vault is not reasonable. The UNIX sync() function suffers the same defect; it schedules a request to harden the data but returns before the operation completes. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
