2006-02-14, 10:36 AM
Not sure whether this belongs in support or wishlist so here's the issue, at what seem like irregular periods throughout the day, like now at 10:25am, GBPVR is hammering the CPU and the disks are rattling.
It's clearly a maintenance function, not the EPG as that is set for 7am.
The problem is that I have two MVP's and it's the instances of the MVP's GBPVR.EXE process that's above normal priority that are doing the work. The MVP's are off so it's not screen update.
The problem is compounded because there are two instances of GBPVR running at above normal priority and they are each clearly trying to do the same task as they each have pretty much 50% CPU utilisation.
If this is database compaction or other scavenging, surely this isn't good to have two instances trying to achieve the same thing at the same time and at a non configurable time of day?
With them both being above normal priority, the machine becomes completely slugged for minutes on end, 10 minutes so far.
So as for the wishlist, if this isn't configurable in any way as it stands, please can a future version check a semaphore too see if the task is already being performed by one instance and can that task be scheduled?
David.
It's clearly a maintenance function, not the EPG as that is set for 7am.
The problem is that I have two MVP's and it's the instances of the MVP's GBPVR.EXE process that's above normal priority that are doing the work. The MVP's are off so it's not screen update.
The problem is compounded because there are two instances of GBPVR running at above normal priority and they are each clearly trying to do the same task as they each have pretty much 50% CPU utilisation.
If this is database compaction or other scavenging, surely this isn't good to have two instances trying to achieve the same thing at the same time and at a non configurable time of day?
With them both being above normal priority, the machine becomes completely slugged for minutes on end, 10 minutes so far.
So as for the wishlist, if this isn't configurable in any way as it stands, please can a future version check a semaphore too see if the task is already being performed by one instance and can that task be scheduled?
David.
