point noted mam.. Thanks for your help !

That is it... nothing more than some login failures...

I concluded Rollback failure from these log entries.

Error log

During undoing of a logged operation in database '', an error occurred at log record ID (xx)

Database was shutdown due to error...

Is the mirroring failing over? No

Are there any useful messages in the error log?

Error log:

4078 transactions rolled forward in database 'database' (9). This is an informational message...

Gail,

we use Ola's scripts for maintenance.

The database doesnt go offline and nor sql service is restarted.

One ques: I understand that Long running rebuild index will take time to roll back....

@Gail, Do you think high VLF count (~750) can be reason for this delay?

I am not fully convinced because I have seen databases with couple of thousands of VLFs...

I run checkdb every weekend, so corruption can be ruled out.

My maintenance plan follows all MS recommendations.

since the issue is not happening right now I cant check log reuse status.

Yes I understand that... Normally we take log backup frequently...

Some time reindexing fills up the entire 100gb drive...

But normally if Tlog is full, process should fail with Tlog full error....

Anybody?

Hi Gail,

What should be qualified as High number of VLFs?

My 300GB database has 750 VLFs (log size 100GB)

I have set growth rate to 1000MB; initial size of 98GB

Question: Why i landed at this thread?

Answer: checkpoint was blocking my differential backup so i wanted to know when did checkpoint start and history.

Alternatively you can try to create trace using TSQL, which provides you more flexibility...

http://msdn.microsoft.com/en-us/library/ms188662.aspx

Can you please explain the question properly?