January 4, 2013 at 5:05 pm
Hi, I have a push transactional replication set up that pushes to a Oracle 11g db. What used to take around 9 to 10 minutes now takes around a hour. There are about 3.5 million rows in this table (well call it table1 for this purpose) on the sql server. The oracle table (well call it table2) also has about 3.5 mill rows. There are only 2 indexs on each table.
I ran a trace to try and find any long duration queries hanging, used sp_who2 to look for locks, cpu, i/o, memory and they all look fine.
We also looked at the Virtual Log Files and they were around 3500 so we shrinked them to around 400 and right now were are going to check if that helped as I type this.
Anything else I am missing your help is much appreciated?
p.s. sorry i originally posted this same topic in another section..
January 4, 2013 at 6:25 pm
Have you also checked the oracle-side? (tracing,locks etc)
January 4, 2013 at 8:51 pm
I was affraid you would say that. I only looked at he cpu, i/o and memory of the Oracle server. I unfortunately am not a Oracle dba. If you have links or quick code snippets that I can run on Oracle to see that would be much appreciated!
Are there any Distributor checks I can do to see if it could be the Distributor that is causing the slowness?
Appreciate the help!
January 7, 2013 at 7:48 am
Haven't time to setup a mssql to oracle replication, however some possible helpfull links.
Perhaps Statspack if installed on the oracle db (small interval)
Dreaded question: anything changed on hardware level? network, storage, db-settings...
January 7, 2013 at 10:23 am
Thanks for the link much apprecaited. I will find if there were any changes thanks.
January 8, 2013 at 3:10 am
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply