What about the row estimates between the servers. Are they the same? You're getting a different memory grant because the optimizer thinks it's moving different amounts of data (and in...

Brandie Tarvin (4/15/2016)

---

Grant Fritchey (4/15/2016)

---

Backups. What are they good for?[/url]

and, more recently:

How to recover from deleting everything, including the backups

I sincerely hope that this data is not vital to your...

Another vote for distributed replay. That's how I've done this sort of testing.

There are absolutely optimizer differences between 2008 & 2012. The cardinality estimator is not different. That's what those traceflags in 2014 are for. It's not applicable to 2012 where the...

It sounds like the process may be running still. It could be blocked by some other process. Have you checked blocking? Use sys.dm_exec_requests to see what processes are actively running,...

Add the index.

A lot of the recommendations for SharePoint databases are not correct. Do what you know is the right thing, deal with Microsoft support if you have to.

However,...

Backups. What are they good for?[/url]

and, more recently:

How to recover from deleting everything, including the backups

I sincerely hope that this data is not vital to your organization, because it's gone.

Or, nuking the offsite backups along with...

BURN IT WITH FIRE!!!!

Jeff Moden (4/14/2016)

---

That's what I thought, as well, but even on the predominately "batch-needs-scans" databases, performance hasn't suffered. Heh... of course, that could be because that code sucks to...

This is a major tuning effort. I'd definitely be able to say more if I could see the execution plan. However, the other comments, JOINs on views, what appears to...

Hopefully you have backups in place. You may need to go to these and be prepared to do a page level restore. This is not an easy operation. Practice several...

I've seen this argument before, and I actually lean towards what Jeff & Brent are advocating... if your queries are primarily seek. If you've tuned the system so you're largely...

I agree with Jeff, it depends.

However, let's say Staging is really Staging, a place to validate deployments prior to production. Then no. I wouldn't have backups at all.

If they're...

You could look to mirroring. The failover to the secondary won't be instant, at all, but it'll almost do what you're looking for.