Depends on how many rows your query affects and how much lock memory SQL has available.

If both are switched off, then you'll probably be getting table locks.

Is S a shared drive? Is it available to Node 2?

Consider using a timestamp (aka rowversion). It's got nothing to do with time despite the name. It's a binary number that changes every time the row is updated is unique...

They affect what logs are allowed on the indexes. SQL can take locks at various granularities depending on the amount of rows affected. It can lock at the row, page,...

You may want to contact Microsoft's customer support people on this. Stack dumps shouldn't be occurring at all, much less every couple min. They will have the tools to read...

And please post in the appropriate forum next time.

Especially if you're trying to corrupt the database.

Log files aren't optional, deleting them should be an absolute last resort and...

Weird...

Are there any entries in the sql error log or windows event log from the same time?

Ok,...

When you get a deadlock graph, post it here and I'm sure someone will help you with it.

What traceflag do you have on to get that output? (1205?)

Try switching whichever that one is off and switching 1222 on. It gives very nice, very detailed output. There's nothing...

The advantage is that the log backup will clear out of the log the entries for the delete that just finished, allowing the space to be reused on the next...

Use the server-side trace functionality. It's what profiler uses behind the scenes and it uses a bunch of sp_trace procedures.

You can get profiler to create a script for you...

They're a big, big organisation, their methodologies and ways of doing things are set and are not subject to change. As long as you don't mind been told exactly how...

Can you do an integrity check on that DB?

DBCC CheckDB (< Database name > ) WITH NO_INFOMSGS