On Wed, 9 Jan 2008 11:31:45 -0500, Thompson, Steve <[EMAIL PROTECTED]> wrote:
>We came to see that we had LSQA (below) growing. > >Now that we know what it is, we can even reproduce this on other z/OS >releases than 1.9 & 1.8. And it appears that we can do it with ANY >program if we can hold it to the same initiator (class selection handles >this nicely). The more complex the program, the more stuff doesn't go >away in LSQA if the program is cancelled. The creep is with IBM's >control blocks. > I think you folks should keep working this via the IBM Support Center, Steve. As I generally believe that while a PMR is open IBMers should not get involved via other mechanisms, such as IBM-MAIN, I won't say much here. However, from the PMR so far I don't see information about anything leaking except ACEE storage. If you've found something else then you should definitely continue working with the Support Center folks to get it explained. As for ACEEs, if a program creates them using RACROUTE, then as documented it has the responsibility to delete them or to use a subpool that the system will clean up automatically. If the program uses the default ACEE subpool of 255 the system won't clean them up on a CANCEL, as CANCEL does not terminate the address space and 255 is a "life of address space" subpool, not a "life of job" subpool. The program could either setup an ESTAE to do the cleanup itself, or it could use a different subpool, either a job-related subpool or a task-related subpool. -- Walt Farrell, CISSP IBM STSM, z/OS Security Design ---------------------------------------------------------------------- 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
