RealHeavyDude
Well-Known Member
OE 11.3.1 64 Bit on Solaris SPARC 10.
I've recently checked the checkpoint interval during the night on our production database where the batch processing in our applications creates ca. 24 GB worth of After Image transaction notes in roughly 6 hours. During that time the checkpoint interval is between 4 - 10 seconds - which is, as far as I know, way too low. Currently the before image cluster size is 65536 which seems already high to me. As the cluster size directly correlates to the checkpoint interval, I think I should increase it, for example double it.
During the rest of the day where are only 6 GB worth of After Image transaction notes generated the checkpoint interval is already somtimes hours.
Does anybody have experience with cluster sizes that large and possible negative impacts?
Thanks in Advance, RealHeavyDude.
I've recently checked the checkpoint interval during the night on our production database where the batch processing in our applications creates ca. 24 GB worth of After Image transaction notes in roughly 6 hours. During that time the checkpoint interval is between 4 - 10 seconds - which is, as far as I know, way too low. Currently the before image cluster size is 65536 which seems already high to me. As the cluster size directly correlates to the checkpoint interval, I think I should increase it, for example double it.
During the rest of the day where are only 6 GB worth of After Image transaction notes generated the checkpoint interval is already somtimes hours.
Does anybody have experience with cluster sizes that large and possible negative impacts?
Thanks in Advance, RealHeavyDude.
Last edited: