Re: Crazy HPL Error....

Neil Truby wrote: > "Tam O'Shanter" <tam@oshanter.com> wrote in message > news:PQCfh.426276$tl2.81137@fe10.news.easynews.com... > >>Greetings, >> >>Muddled my way through unloading a large IFX 10UC4 database and am trying >>to work through the load process. >> >>HPL is really amazing at unloading large amounts of data an I'm thoroughly >>impressed.. >> >>When calling onpload to execute a load job I get the following: >> >>plsig_segv(): received SIGSEGV >>13:35:03 pl_main.c, line 593, thread 13, proc id 18673, Segmentation >>Violation. >>plsig_segv(): received SIGSEGV >>13:35:03 pl_main.c, line 593, thread 1612725740, proc id 18835,>>Segmentation Violation. >>plsig_segv(): received SIGSEGV >>plsig_segv(): received SIGSEGV >>3: 3: No such file or directory >>13:35:03 semop: errno = 22> > > An error 22 is a shortage of some OS resource. Are you sure you have the > kernel settings as per the release notes. > Also, which OS? > > Actually errno 22 is: Bad argument to a system call. Usually it means you passed a msg queue id, shared memory id, or semaphore id that doesn't exist to the appropriate system call to access that resource. Since the IDS code is fairly shaken out here, it's usually the result of a VP crashing causing some system object to go out of scope and be deleted causing subsequent access by another VP to get the errno 22. In this case, the SEGV crashed the task causing the errno 3 and errno 22 in another process. Symptom not cause. Art S. Kagel