Saturday, November 20, 2010

Oracle Redo log file size

I was trying to delete a number of rows (1-2M) from a table on the new production server last night and found the operation was taking far too long. Although there are delete optimization techniques, like using temp tables, it certainly wasn't anywhere near enough records to have to implement something like that.

A little digging showed that the operation was blocking on "log file switch (checkpoint incomplete)". Resizing the log files did the trick.

No comments: