Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.
Moderators: chulett , rschirm , roy
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri May 29, 2009 9:02 am
I am tryin to load data using ODBC stage. I am wusing Upsert mode and what i realise is its extremely slow some times close to 37 rows per second is there any way i can improve the speed
Thanks
Shyam
chulett
Charter Member
Posts: 43085 Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO
Post
by chulett » Fri May 29, 2009 9:09 am
Is 'upsert' really necessary? Can you use a native Enterprise stage instead? Would help to mention the target db as well.
-craig
"You can never have too many knives" -- Logan Nine Fingers
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri May 29, 2009 9:20 am
Yep, I need to capture rejects so i am using upsert. The target is ORACLE 11g Since we dont have drivers we are using ODBC
Shyam
chulett
Charter Member
Posts: 43085 Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO
Post
by chulett » Fri May 29, 2009 9:25 am
Are you doing the upserts in the 'right' order, meaning whichever action has the best chance of succeeding first? For any updates, are you updating on indexed columns?
-craig
"You can never have too many knives" -- Logan Nine Fingers
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri May 29, 2009 10:07 am
Actually all are inserts because i need to capture rejects. I am using upsert, As reject option is not available in Write
Shyam
chulett
Charter Member
Posts: 43085 Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO
Post
by chulett » Fri May 29, 2009 10:11 am
Okay... Array Size?
-craig
"You can never have too many knives" -- Logan Nine Fingers
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri May 29, 2009 10:22 am
60000
chulett
Charter Member
Posts: 43085 Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO
Post
by chulett » Fri May 29, 2009 10:46 am
Seems... high. What is your ARL, average record length? Does it help if you drop it down to a smaller number, say like 500 or 1000?
-craig
"You can never have too many knives" -- Logan Nine Fingers
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri May 29, 2009 12:36 pm
Its barely 7-8 columns which amounts to 100 -150 bytes
chulett
Charter Member
Posts: 43085 Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO
Post
by chulett » Fri May 29, 2009 1:04 pm
Still... did you try other values?
-craig
"You can never have too many knives" -- Logan Nine Fingers
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri Jun 19, 2009 8:15 am
Yeah,
The record length is really small. Its extremely slow ...I have no idea what to do..please suggest to improve the performance
Shyam
miwinter
Participant
Posts: 396 Joined: Thu Jun 22, 2006 7:00 am
Location: England, UK
Post
by miwinter » Fri Jun 19, 2009 8:21 am
Is obtaining the Oracle drivers so much of a hurdle? Strikes me you'd be better off doing what Craig originally suggested and use a native Enterprise stage as opposed to ODBC, if performance is your 'driver' (sorry!) :D
Though you say "all are inserts", is that just by nature of your processing, or have you actually set "insert only" or "insert then update" on your upsert mode?
Mark Winter
<i>Nothing appeases a troubled mind more than <b>good</b> music</i>
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri Jun 19, 2009 9:12 am
Insert Only
sgubba
Participant
Posts: 30 Joined: Wed Apr 16, 2008 11:06 am
Post
by sgubba » Fri Jun 19, 2009 9:14 am
I am trying to insert in to Oracle 11g. Thts the reason we are using ODBC
miwinter
Participant
Posts: 396 Joined: Thu Jun 22, 2006 7:00 am
Location: England, UK
Post
by miwinter » Fri Jun 19, 2009 9:20 am
So there are no drivers for Oracle 11g then?
Mark Winter
<i>Nothing appeases a troubled mind more than <b>good</b> music</i>