Re: Database Replication/Checkpoint Delays

If your checkpoint time had increased after replication (HDR) started, you can add ON_RECVRV_THREADS in configuration file on secondary server. On primary, LRU_MAX_DIRTY and LRU_MIN_DIRTY can be set as low as 2 and 1 (which maximizes flashing between checkpoints). Arun Shastry wrote: > > Try to increase your cleaner threads and/or AIO VP's. They don't put much > of an overhead on the system, but might help speed-up checkpoints. Also, > check your LRU_MAX_DIRTY and LRU_MIN_DIRTY parameters > and lower them if you have to. See the ratio of LRU writes vs. checkpoint > writes in your onstat -F output. Also, see if any of your chunks are the > I/O bottlenecks with onstat -g iof. > > When all said and done, if you still have checkpoint delays, you might want to > > look into restructuring your table layous for even I/O distribution. > > Good luck, > Arun > > Greg McGregor wrote: > > > Hi, > > > > My company is operating Informix Online Dynamic Server 7.14 on 2 x HP-UX = > > 9.04 with data replication. > > We are experiencing checkpoint delays on the primary server, every now = > > and then, and I would like to > > find out whether anyone else has experienced this phenomenon while = > > running data replication. > > > > <snip>