July 4, 2011 at 6:51 am
During the process of creating a large data flow process I missed a field.
From Source to Destination I have 50 flows comprising of Sorts, Merges, Splits and Derive Field processes to go through.
Is there an alternative to opening every flow and checking the field to pass through??
July 4, 2011 at 7:42 am
I'm afraid not, SSIS is strongly typed about its metadata.
On a sidenote: isn't a dataflow with 50 flows a little too big? Can't it be split into smaller packages?
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
July 4, 2011 at 9:57 am
Koen Verbeeck (7/4/2011)
I'm afraid not, SSIS is strongly typed about its metadata.On a sidenote: isn't a dataflow with 50 flows a little too big? Can't it be split into smaller packages?
Fair enough.
1 Data Flow with 23 Sorts & 23 Merges, 1 Split and 2 Derives.
Combining lots of Reference data with several Data tables so not sure how I'd split it down across multiple flows
July 4, 2011 at 10:06 am
Michaelro (7/4/2011)
Koen Verbeeck (7/4/2011)
I'm afraid not, SSIS is strongly typed about its metadata.On a sidenote: isn't a dataflow with 50 flows a little too big? Can't it be split into smaller packages?
Fair enough.
1 Data Flow with 23 Sorts & 23 Merges, 1 Split and 2 Derives.
Combining lots of Reference data with several Data tables so not sure how I'd split it down across multiple flows
Eek, that sounds nasty. What are the sources and destinations of the data? (File / SQL Server / spreadsheet) etc
The absence of evidence is not evidence of absence.
Martin Rees
You can lead a horse to water, but a pencil must be lead.
Stan Laurel
July 5, 2011 at 3:33 am
OMG !! .. Are you running it on a super computer !! .. just kiddin...
July 5, 2011 at 12:39 pm
Koen Verbeeck (7/4/2011)
I'm afraid not, SSIS is strongly typed about its metadata.On a sidenote: isn't a dataflow with 50 flows a little too big? Can't it be split into smaller packages?
Not only is SSIS strongly typed it's also not at all proactive about cascading changes you make to column metadata through a dataflow.
And yes 50 dataflows in one dataflow seems scary.
July 5, 2011 at 11:25 pm
ZZartin (7/5/2011)
Not only is SSIS strongly typed it's also not at all proactive about cascading changes you make to column metadata through a dataflow.
Luckily, this will change in the next version of SQL Server.
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
July 6, 2011 at 2:24 am
ZZartin (7/5/2011)
Koen Verbeeck (7/4/2011)
On a sidenote: isn't a dataflow with 50 flows a little too big? Can't it be split into smaller packages?And yes 50 dataflows in one dataflow seems scary.
Really?
You need to sort data before it's merged so that's 5 Flows per Merge (2 Sources + 2 Sorts + 1 Merge)
My Source data contains all Lkup Codes and I'm adding in the Descriptions to create a Single Table.
July 6, 2011 at 11:02 am
I think a terminology mistake is causing these replies.
You don't have "50 flows" in your data flow, you have 50 components.
50 "flows" in one Data Flow Task is downright bad architecture. 50 components is not even close to stressful.
Todd McDermid - SQL Server MVP, MCTS (SQL 08 BI), MCSD.Net
My Blog - Dimension Merge SCD Component for SSIS - SSIS Community Tasks and Components
July 7, 2011 at 1:57 am
Todd McDermid (7/6/2011)
I think a terminology mistake is causing these replies.You don't have "50 flows" in your data flow, you have 50 components.
50 "flows" in one Data Flow Task is downright bad architecture. 50 components is not even close to stressful.
Sorry - Bit green with SSIS so haven't got the terminology right yet obviously :unsure: 🙂
Thanks
July 7, 2011 at 2:54 am
No problem, just trying to save you sone grief!
good luck with SSIS, and have fun...
Todd McDermid - SQL Server MVP, MCTS (SQL 08 BI), MCSD.Net
My Blog - Dimension Merge SCD Component for SSIS - SSIS Community Tasks and Components
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply