#D14%DDATHD21.BITNET@WISCVM.ARPA (09/09/85)
We have a problems with the F11XQP. One userprocess is consuming a lot of cputime and most other processes are waiting (LEF or LEFO). The VMS must be stopped with @CRASH at the console. Looking into the crash dump I found a lot of processes in the ACP disk wait queue. The first process of the queue is the process which is consuming the whole cputime. This problem occured more than five times on different VAXes. All people with file accesses are waiting in the ACP queue. Other programs that don't need file access (i.e. SET HOST, PHONE, MONITOR) are still working. Any idea ? Reinhard Goeth @ Technical University Darmstadt, W.Germany, Europe (Beware of the number-sign. It's part of my userid !!!)
linscomb@NGP.UTEXAS.EDU (Thomas J. Linscomb) (09/09/85)
> Subject: F11XQP (V4.1) problem > We have a problems with the F11XQP. > One userprocess is consuming a lot of cputime and most other processes > are waiting (LEF or LEFO). The VMS must be stopped with @CRASH at > the console. Looking into the crash dump I found a lot of processes > in the ACP disk wait queue. The first process of the queue is the > process which is consuming the whole cputime. This problem occured > more than five times on different VAXes. All people with file accesses > are waiting in the ACP queue. Other programs that don't need file > access (i.e. SET HOST, PHONE, MONITOR) are still working. > Reinhard Goeth @ Technical University Darmstadt, W.Germany, Europe We have also had this problem or at least the symtoms as you describe them. The only machine it has been occuring on though is one that has multiple types of disk drives. Ours has an RM80 and two RP07s (with a user mountable RM03). None of the other sites with only one disk type have had this problem. This information is also confirmed with another site manager in town. The solution we opted for is to use the F11XQP from a v4.2 system. This has only been in for about a week so I can not give you any results. Be aware that we were warned that this may introduce other problems. I know this is not much help, but at least you know you are not alone. --Thomas aka linscomb@ngp.UTEXAS.EDU Advanced Graphics Laboratory The University of Texas Computation Center