SLES10 - network failure on DataGuard log streams

Some time ago I found a strange problem with Oracle @ SLES10 (used SP1 beta
those time).
- Under a heavy load;
- With Oracle Data Guard configured;
- Primary database reported Async IO error on archive logs stream, and so
lost syncronization with standby
database (can't ship logs anymore);
- It recovered in time, because of FAL feature of the DataGuard (standby
pull absent logs and syncronize);

Configuration used SLES10 on one side and UL4.4 on another side, and problem
occur only when primary DB was on SLES10.

Now I retested the same configuration with 2 x SLES9 SP3 servers, and
everything works fine.

So I can conclude, that there is a bug in Oracle implementation on Linux
SLES10 (or in SLES10 related to the network IO, or
in Oracle sysctl setting for SLES10 - have not time to investigate).

Did anyone seen anything like this on SLES10?


-- 
To unsubscribe, email: [email protected]
For additional commands, email: [email protected]
Please see http://www.suse.com/oracle/ before posting