Viewing 13 posts - 1 through 13 (of 13 total)
Here is a question, if you select 5mm or selecting 10Kx500 won't the time it takes to do the select equal in any case?
If so what is the point in...
August 20, 2010 at 2:25 pm
Ok, thanks to everyone’s help!
I’ll just continue trying to speed it up
August 9, 2010 at 12:25 pm
Well I periodically run maintenace scripts on it (like re-indexing etc)
My largest involved table is:
2.440Gb in dataspace, 1.213Gb in index space, 19 307 075 total rows
Second largest involved table:
881.711 MB...
August 9, 2010 at 7:35 am
Ok, we are using RAID1 6x500Gb into 3x500Gb drives (C,D,E)
TempDB on C, mdf on D and ldf on E (file growth set to 500mb)
External SAN as backup, just for interest...
August 8, 2010 at 12:45 pm
Nope no triggers or any additional processing on the tables.
August 6, 2010 at 4:04 pm
I used SQL Profiler and the log is allowed for autogrowth by 10% limited to 10Gb, last time I checked.
August 6, 2010 at 3:47 pm
Well, let me list some more stats:
average IO write ops 4~47 / sec, avg 188434 bytes/ sec (running raid thus using averages) , half way through it starts to peak...
August 6, 2010 at 3:30 pm
Thanks for the feedback guys, but unfortunately I only upload 400K entries at a time never more using 1 Transaction and commit at end.
Also Isolation level are set to minimal...
August 6, 2010 at 4:51 am
The script looks like this:
INSERT INTO [OPR_SUPER_BOX] (SBox_No,Job_ID) VALUES ('000000000000000',152)
INSERT INTO [OPR_BOX] (Box_No,Job_ID,SBox_No) VALUES ('000000000000000000000',152,'0000000000000000')
INSERT INTO [OPR_BATCHES] (Batch_No,Box_No,Job_ID) VALUES ('00000000000000000000000','00000000000000000000000',152)
INSERT INTO [OPR_SERIALS] (Serial,Job_ID,Batch_No,Exp_Date,Misc,SCID)VALUES('0000000000000',152,'00000000000000000000000','','','')
INSERT INTO [OPR_PINS] (Serial,Job_ID,Pin_Num,Pin_Set,Pin_Serial) VALUES ('000000000000',152,'0000000000000000000000000000000000000000000000000000000000000000',1,'000000000000')
INSERT INTO [OPR_PINS]...
August 5, 2010 at 2:47 pm
We did use batch processing originally, can't remember why, but found that row-by-row approach worked more efficiently back then when we were still using SQL2000 (now SQL 2005 Std).
We run...
August 4, 2010 at 1:56 pm
I checked the script durations and went like this:
The first 10min duration 0, at 10min it drops to 35, at 20min to 110 and after 30min it goes to 500...
August 4, 2010 at 10:36 am
Makes sense, I'll take a look at script execution times.
Thanks
August 4, 2010 at 8:47 am
Well let’s elaborate the situation:
Technologies used:
Server is a Quad Xeon 2.45Ghz (or something like that) 16Gb Ram, 1.5Tb total storage divided into 4 drives 1 for Windows, 1 for mdf...
August 3, 2010 at 12:11 pm
Viewing 13 posts - 1 through 13 (of 13 total)