Re: Data Replication Delay

I experience this same problem on two HP9000 H60 running 7.12.UC1. Under certain conditions (depends on your application), DR can hang. To be specific, we use our secondary server as an online-querying engine for reports/users, and when long queries were performed, DR's priority was low, so queries took precedence over the replication of data. When you do a "onstat -g dri" on the primary what messages do you get? Do you get a "checkpoint expected" on the primary or secondary? Are you using your secondary server as a querying engine or is it just sitting there? The solution for us, was that we received a "patched" engine from Informix (this was after months of demonstrating the problems to them). Now, data replication works as expected, but our performance has sufferred. We are currently looking to 7.2, to see if our "patch" was included in that release. > > We have two of the above systems, one is the primary server and the > other is the secondary server. Our checkpoints occur ever 5 minutes > and take 0 seconds (occasionally 1 second). > > We are running FourGen Accounting compiled 4GL code on the production > system and using Arpeggio for our reporting tool via ODBC (Informix > Connect/Visigenic). > > The problem is that every once in a while the production system will > hang. After analyzing the logs we have several messages, it starts on > the secondary system with a "DR: ping timeout" message in the log > file. (We can ping the systems and there is not problems, the timeout > variable is set to 60 seconds in the config file) Then on the primary > system, the next checkpoint will take between 100 to 1500 seconds. > During this time the system seems locked. The system does eventually > resync and works as normal. > > My question is: Has anyone experienced this problem?