I was first asked to look at this problem as a database restore was getting an error when attempting to restore the archived logs: archived…
A database that had been migrated to AWS went live and was hitting serious performance problems three times every weekday. I was asked to investigate…
Why would performing an index fast full scan 430 times, reading 38MB and getting back 297 rows cause a query to take almost an hour?
In fact I could find no relevant errors in any of the log files…
A quick look at the relevant statistics showed why there was such a significant difference
However, two days later whilst still eating my breakfast I was asked to join a conference call about a bigger problem
The errors in the alert log were far more frequent than the database hangs, so clearly there was another factor involved …
There were all kinds of errors in the log, but the one of most concern was …
This is more by way of amusement than an actual investigation. I had been asked to monitor a test of a new piece of application…
This is a rather quick investigation, but it shows some interesting Oracle behaviour. I had written a PL/SQL package that manages range partitioned tables: adding…