Re: checkpoint duration

Try these: 1. Increase the size of the physical log to eliminate checkpoints triggered by that being 75% full 2. Increase the NUMAIOVPS until you see the IO operations performed by the AIO VP's in onstat -g iov gradually reducing (usually one to two orders of magnitude)3. You probably want to reduce your checkpoint interval also. 4. Bump down the LRU_MAX_DIRTY and LRU_MIN_DIRTY to allow LRU writes in the background. 5. If your platform supports Kernel AIO, enable it. Then, see if you get decent performance. Things to do after that for finer tuning: 6. Reduce the LRU's to whatever your number of CPU VP's is Hope that helps, Arun Willie L. Meeks wrote: > I have a serious problem with checkpoint duration which cause my posting > process to almost double in duration because of the amount of time wasted > on checkpoints. Here is a a sample of the duration take directly from the > online log file. > > <snip> > Regards, > > Willie Meeks