You’re offline. This is a read only version of the page.
Hello,
I've run into a situation where I need to update MR from RU3 to CU11.
Installing the updated software is not an issue until I get to the point where I run the update database procedures.
The initial size of the MR database is already large at around 40gb.
I've check and only 18mb of space is unused in the database.
The database is Legacy, and when I attempt to update it, the LDF grows to around 200 GB, before the server crashes. The MDF is sitting around 80 GB at this particular point.
I'm trying one last time to dedicate a new server with 1 TB of disc space to see if the crash is related to the drive running out of space completely.
Does anyone know why the LDF is growing so large, or a way to prevent / calculate how large the MDF and LDF will grow during the update?
Thanks,
*This post is locked for comments
I opened a case with Microsoft. Basically their answer is that the database structure changes so much from RU3 to CU11 that the upgrade process is copying the whole 40GB database to several temp tables to accomplish the upgrade. The answer was, 400 GB os space was needed for the update... period.
They were, however, kind enough to offer me some scripts to run to free up space, and in the end after running them and running a shrink database task, the LDF is gone, and the MDF is 58 GB (18 GB larger than when i started). Much better than 90 GB.
One thing I didn't read right in your original post is that the DB is at 40gb w/ 18mb of space. Try expanding the size of the actual database by a couple hundred megs and see what happens. Depending on how you have your database setup this would be the cause of your LDF blowing up. SQL is just running transaction after transaction against the database but it very well might be waiting on the OS to allow SQL to autogrow your database. If you have a 10% auto growth SQL will grow your database by 4gb when out of space. Also you can look at Instant File Initialization to reduce the time SQL has to wait to grow.
Do not shrink a database as you will create fragmented data. There could be several reasons why your log file grows when you update the MR database most likely something is causing the transactions from being fully committed to the database allowing the checkpoint to be made and the log to automatically truncate and continue.
This TechNet article outlines in detail how the process works. technet.microsoft.com/.../2009.02.logging.aspx
You might try running a trace on it to see what happens or check the database waitstats dmv to see what SQL is waiting on also there are some scripts that will show you process that are blocking or deadlocking which would lead to what you mentioned above.
Hi Jonathan,
i hope you have selected shrink of the MR database in the task.
Oh yeah, sorry, I verified the database was in simple recovery model already. (Just to be sure, I did it once more just now)
Hi Jonathan,
This is due to the database is set to full recovery please change the same to simple. you can do this change by going to SQL sever management studio.
Stay up to date on forum activity by subscribing. You can also customize your in-app and email Notification settings across all subscriptions.
#1
André Arnaud de Cal...
291,240
Super User 2024 Season 2
#2
Martin Dráb
230,149
Most Valuable Professional
#3
nmaenpaa
101,156
Share
Report
All responses (
Answers (