Viewing 15 posts - 4,336 through 4,350 (of 7,502 total)
the scripts just checks if these startup parameters exist for your sqlserver instance,
and adds them if needed.
This way, the deadlock notifications in the sqlserver errorlog are always logged, even if...
November 7, 2008 at 12:41 am
am (11/6/2008)
ALZDBA (11/6/2008)
Is this a migrated database from sql2000 ?if so:
- rebuild all indexes
- sp_updatestatistics
- dbcc updateusage(0) with count_rows
and if you can ... dblevel to 90 !
can you please...
November 6, 2008 at 11:57 pm
GilaMonster (11/6/2008)
dbcc traceon(1204, 1222, 3605, -1)
3605 isn't required for deadlock graphs. Personally I find 1204 to be unnecessary if 1222...
November 6, 2008 at 11:48 pm
Is this a migrated database from sql2000 ?
if so:
- rebuild all indexes
- sp_updatestatistics
- dbcc updateusage(0) with count_rows
and if you can ... dblevel to 90 !
November 6, 2008 at 1:56 pm
add the following startup parameters to your sql instance.
This way you'll see in the sqlserver errorlog what statements are suffering
the conflict(s).
--dynamic activaton
dbcc traceon(1204, 1222, 3605, -1)
declare @MaxSQLArg varchar(50)
declare @NewSQLArg...
November 6, 2008 at 1:53 pm
can you run sp_updatestistics and dbcc updateusage(0) with count_rows ?
This should get you fresh stats ...
November 6, 2008 at 1:32 pm
starting a SQLTrace will also help to determine the source of your deletes.
November 6, 2008 at 1:20 pm
bring on the caffeine.
We have a volume licensed Dev Edtn.
Also keep in mind, if you don't supply a key, it will install a Ent edtn for 140 days.
November 6, 2008 at 10:56 am
You can off course create a trigger for delete on that table to perform a rollback for delete statements and raise en error stating all connection info.
November 6, 2008 at 10:55 am
did you check out backup compression softwares ?
They typically reduce the size to +/- 20%
November 6, 2008 at 10:20 am
there is off course the undocumented :crazy: [sp_MSforeachtable].
Keep in mind the order of execution if you have DRI.
(or repeat more than once)
If you want to avoid log overhead, try to...
November 6, 2008 at 10:17 am
use an insert/update trigger to check the condition and rollback if invalid.
November 6, 2008 at 10:09 am
Before you start the restore to recover PIT, you should still backup the current log of the db !
After that, you can restore your previous full backup ( PIT,...
November 6, 2008 at 7:55 am
btw
did you check the plan using a @variable in stead of the hard coded
WHERE vh.VendorID = 14
Maybe vendorid 14 just generates a less optimal (non...
November 6, 2008 at 7:37 am
A trace file does not switch by day.
run
SELECT * FROM ::fn_trace_getinfo(default)
If you don't see an active trace, someone did disable the default trace for your instance.
There are...
November 6, 2008 at 6:34 am
Viewing 15 posts - 4,336 through 4,350 (of 7,502 total)