If you don't know exactly what kind of deadlock you are getting set up a Server Side Trace for Deadlock graph event and find out.

In your situation the deadlock may...

GSquared (7/19/2012)

---

seregak76 (7/19/2012)

---

GSquared (7/16/2012)

---

"Runnable" isn't about locks. It's about resource-availability. CPU, et al. It's waiting for a scheduler to be available to pick it up and actually...

Thanks, I looked at...

GSquared (7/16/2012)

---

"Runnable" isn't about locks. It's about resource-availability. CPU, et al. It's waiting for a scheduler to be available to pick it up and actually get it...

Thanks for clarification, I got it.

This SPID get stuck forever unless i rebuild statistics and this is hard to explain if there is no CPU contention.

I need to see what...

Thanks GSquared. Now my question is why a process gets stuck in status runnable for a long time? It does not wait for any locks to be released.

Your change to use DATEDIFF function will make your qury miss the index on adddatetime.

I totally agree with Jared about changing the purge method. You don't have many choices when dealing with such big volumes. The good news is that your clustered index is...

Hi,

It seems to me that you passed the threshold where GHOST CLEANUP process can keep up with your deletion rate. Can this be the case here?

Thanks everybody!

I got it sorted out. Some times the ' leaf_delete_count' does not match amount of deletes. The rest of deletes are reflected in 'leaf_ghost_count' so I can...

Thanks!

I increased the amount of rows in my test case and still no luck. Any other ideas?

SET NOCOUNT ON

IF OBJECT_ID('t1', 'U') IS NOT NULL DROP TABLE t1

CREATE TABLE t1 (a...

Hi!

I run the trace for Dedlock graph that can give you all information about the deadlocks on server. I start the trace when SQL Server Agent starts and I never...

@sgtwilko

I've implemented updlock as you suggest in your comment and I encountered duplicates that occur very rarely, and not every time when I run tests with many concurrent...

it is not completely true. Although it says in BOL: "At the beginning of a connection, this setting has a value of -1. After it is changed, the new setting...