scraze (2/9/2009)

---

One thing though - I am still wondering a little about whether my original suggestion was possible, using something like:

ALTER TRIGGER [Logging].[MasterLogTrigger]

ON [Logging].[MasterLog]

...

The slide said 260,000 users per day, even though the speaker said 260.

I don't want to derail the topic, but with regard to triggers not rolling back if raiserror is used, BOL (http://msdn.microsoft.com/en-us/library/ms178592.aspx) states

When RAISERROR is run with a severity of 11...

Well, it costs quite a bit less! But I'm not sure that was quite the...

Too complicated for me to really take it all in, but a couple of thoughts:

1) the "with recompile" statement has been commented out - does this mean you tried it...

Bruce W Cassidy (2/8/2009)

---

Several issues come into play here. The first is the "user proxy" issue. Namely, in n-tier applications, the "user" who connects to the database may...

Talk to the user's manager about the request, the reasons why it's a bad idea, the alternatives, and any applicable legal/standards requirements. If they still want you to proceed...

You may have to play with the restore options a little to fit your exact requirements, but if it is a fixed file name and restoring over an existing database...

Cherie, your problem may be unrelated. Windows groups cannot be assigned a default schema inside SQL. If the application is not referencing objects with a 2 part identifier...

The only way you can force them to use the correct query is to create procs and deny them read access to that table - it probably won't be as...

If the table has a timestamp you could index it and have them include it in all queries - presumably they only want to retrieve recent information. If they...

Yes, it is possible to shrink the log file in a log shipping scenario. But the root cause of the problem is that you are reorganizing all your indexes...

One other option to consider: FTP. As long as you can provide a machine that is reliably connected to the internet and can be suitably secured, this is pretty...

I would guess that you will also get questions about backup / recovery processes - think filegroup backups, striping backups over multiple drives, possibly snapshots and snapshot restores etc. ...

You don't have an actual plan do you? It's possible that the stats are off somewhere, and with a plan this complex a small difference in one node could...