Adaptive Server 15.0.2 adds information to the following monitoring tables.
monOpenPartitionActivity reports information about the events for each partition in the housekeeper garbage collection queues. The columns added are:
HkgcRequests – reports the total number of events queued for the partition.
A large value reported by HkgcRequests implies the sytem is generating lot of garbage for that particular partition or object.
HkgcPending – reports the number of pending events for the partition.
Large value reported by HkgcPending implies a lot of garbage is yet to be collected, although the housekeeper will clean the garbage up. However, if Adaptive Server is rebooted, all the entries in the housekeeper queues are lost and the garbage from those page is not collected when you next start Adaptive Server.
HkgcOverflow – reports the number of partition events that overflowed
A large value reported by HkgcOverflow implies the housekeeper queues are getting full and the garbage that is generated will not be cleaned because the housekeep can not schedule the job.
These columns report on per-partition basis, and give an idea of the amount of garbage generated in each partition for different objects.
monOpenObjectActivity reports information about the events for each object in the housekeeper garbage collection queues. The columns added are:
HkgcRequests – reports the total number of events queued for the object.
A large value reported by HkgcRequests implies the sytem is generating lot of garbage for that particular partition or object.
HkgcPending – reports the number of pending events for the object.
Large value reported by HkgcPending implies a lot of garbage is yet to be collected, although the housekeeper will clean the garbage up. However, if Adaptive Server is rebooted, all the entries in the housekeeper queues are lost and the garbage from those page is not collected when you next start Adaptive Server.
HkgcOverflow – reports the number of object events that overflowed
A large value reported by HkgcOverflow implies the housekeeper queues are getting full and the garbage that is generated will not be cleaned because the housekeep can not schedule the job.
monDeadLock reports the locking information (filename and line number) collected from the lock record deadlock chain. The columns added are:
HeldSrcCodeID – filename and line number where Adaptive Server requests the holding lock
WaitSrcCodeID – filename and line where Adaptive Server requests the waiting lock.
monLocks adds the SourceCodeID column, which reports locking information recorded in the lock record structure:
SourceCodeID – Filename and line at which the lock is requested
