RE: arch slow and standby.
- From: igor.polishchuk@xxxxxxxxxxxxx
- To: oracle-l@xxxxxxxxxxxxx
- Date: Tue, 10 Apr 2007 11:30:02 -0700
Ankur, On 9i, when a network connection is not dead but just slow, you may have a problem. I've experienced problems with a live database. It actually hanged under heavy load. I may not remember exact details, but it was definitely a problem because archiver waited until remote archive would be created. To fix this situation, look at hidden parameter _log_archive_callout and set it to 'LOCAL_FIRST=TRUE'. If you make a search on metalink, you'll find an article describing exactly mentioned above behavior and recommending the mention parameter setup. I did not have a chance to really test it on a slow network myself yet. _____ From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of agodambe@xxxxxxxxxx Sent: Friday, April 06, 2007 1:21 PM To: oracle-l@xxxxxxxxxxxxx Subject: arch slow and standby. Hi, In a 9i managed standby database with primary and secondary in maximum performance mode if arch is not able to archive to remote destination due to n/w issues etc, it will retry archiving to remote destination depending upon REOPEN and MAX_FAILURES parameter. But what if the n/w is slow (not dead and arch is actively archiving redo log) and by the time it's archiving to the remote destination all the log groups on primary are filled and lgwr comes back to the group being archived currently for reuse. Will the database hang? Or it will stop archiving to remote with error message somewhere in dictionary? I tried simulating but since I have the test databases on lan couldn't simulate. I think it will wait till it's archived to the remote. Anyone who has faced similar situation or knows what might happen? Thanks, Ankur