Re: Error when trying to access output queue

On 2/10/11 8:12 AM, Albert York wrote:

The QEZJOBLOG output queue has over 13,000 entries in it but when I
try to display it or clear it I get the message Permanent machine
error X'0207'. Internal dump identifier (ID) 0300C4C6.

This is a V4R5 system. Is there any way to delete the queue?

That is a LIC Function Check which produces a Vlog VL02000207 and the identifier suggests PRTINTDTA *DMP DMPID(0300C4C6) will spool the VLog; though often other nearby VLogs will be relevant for the failure. That error could mean just about anything, but I typically saw those as the side effect of past disk errors; e.g. a page of zeroes is passed to the OS from the LIC SM for a page that can not be accessed, even for the instruction stream of a program. In this case the error may be with one of the internal spool block indexes for tracking the spools [by output queue]. Investigating the cause of the error might lead to means to circumvent; of course having such an old release may not be of much luck in finding assistance to review the VLogs and PALs.

Whatever is encountered by the WRKOUTQ may be circumvented by CLROUTQ, DLTSPLF, CHGSPLFA, or WRKSPLF. For example, WRKSPLF *ALL OUTPUT(*PRINT) ASTLVL(*INTERMED); if functional, the output would show the details of the SPLF on the QEZJOBLOG *OUTQ. Similarly, spool list API requests [probably other than by OUTQ] might enable bypassing the errors.

Note that there are one or two spool control block cleanup programs that may be able to detect and correct a problem with a spool index. A harsher and more thorough method involves cleaning up all spools and incomplete jobs at IPL, I think optionally preceded by another heavy [and long running] operation.

Regards, Chuck