Are you sure that the SAN LUN was present and available at the time the SQL Service started? Could it have been that the drive only became available afterwards?

Ok, it's literally an internal per-database flag that says if there's an existing full backup for the purposes of running a diff. If the flag isn't set, diff's will fail....

Link to article please? Not sure what it's talking about.

Just bear in mind that unless you did the backup with the CheckSum option, a successful RESTORE VERIFYONLY does not guarantee that the backup is restorable. As far as I'm...

Good to hear it's fixed. If you haven't already, get the server admins to run some checks on the disk, make sure there are no lingering problems

Looks like the shutdown was requested by someone.

Basically, someone logged in as sa ran this

SHUTDOWN WITH NOWAIT

Is this...

That doesn't look like a SQL Server error log. It looks like the SQL Agent error log, the reference to idle CPU condition is something that'll only appear in agent....

Easiest way to get the execution plans is simply to run the queries manually. Question is whether or not they'll generate identical plans to the ones the app runs. If...

It doesn't suggest anything about the index maintenance. It does possibly indicate stale stats, but they shouldn't get stale that fast. How big's the table, how fast are you inserting...

But surely 02/12/09 7:00AM is bigger than 01/12/2009 23:30PM

Alter Index

Sysprocesses is deprecated and should not be used in new development. It will be removed in a...

So there's only 9 MB free in a 51 GB database. Definitely not worth a shrink.

If you rebuild all the indexes on the table that you deleted rows from,...

Sounds like a parameter sniffing problem, not a statistics problem. The stats update would fix things, because it forces plans to recompile.

Can you post the stored procs that are used...

chris.worthington (12/1/2009)

---

2). From the majority of postings I've read, the recommendation is to have AUTOSHRINK switched off and replaced by scheduled DBCC SHRINK commands instead. Is that a fair representation...