I am currently testing our upgrade process in a test environment. I know that since our VM's are not as highly spec'd as our production environment, we should expect some decreased performance. However, after scouring the community site to determine why timeouts were occurring, how to possibly avoid them, and how to possibly improve the completion time, I am still at 10 hours or 36656 seconds to complete the upgrade.
Changes applied:
- Increase timeout to 600,000, successful(support suggested increasing to 999,999). Attempted increase from 300 to 600, failed, timeout exceeded. Attempted increase from 600 to 1200, failed, timeout exceeded. Attempted increase from 1200 to 12000, failed, timeout exceeded.
- Changed database recovery mode from Full to Simple and trimmed the Transaction Log
USE [DatabaseName];
GO
--(1) first, set recovery mode to SIMPLE
--(2) run the following to reduce log to 200mb
CHECKPOINT;
GO
CHECKPOINT;-- run twice to ensure file wrap-around
GO
DBCC SHRINKFILE(OOTB_60_log, 200); -- this uses the "logical name," not the actual log file name
GO
3. I did not Check user-defined Indexes and Statistics dependent on converted columns, as I did not get a failure due to this.
I debated check marking Unicode prior to running any upgrade, but realized it was already too late. Any opinions or additional information would be greatly appreciated.
Thank you,
Mike