Just my 2 cents worth..

I agree with the other posters, you cannot feasibly use 5M rows in a web page efficiently, unless you are consuming that using some web scraping...

OK, I see where the confusion has arisen,

ProcessTask:WriteToLog(Start) and ProcessTask:WriteToLog(End) would be the same object with two process flow arrows into it. The writing of the log would happen...

I want one PT called multiple times. This way if I need to change it's configuration I only have to do it once.

Maybe I am approaching this the wrong...

No.., maybe I diodn't explain it clearly enough.

I can write the Process task to build the argument string from variables, What I was struggling to do was to define...

The problem with that is that every tme you make changes to the script in DEV you have to modify it again once it has passed UAT so that you...

Ermmm,

By adding a Process Task object to the start and end of the process Flow??

Or am I missing something?

[font="Courier New"]ProcessTask:WriteToLog(Start) -> SQLTask:RunScript -> FlatFileTask:DumpTableToCSV -> FileSystemTask:CopyFile -> ProcessTask:WriteToLog(End)[/font]

Joe,

It seems you are getting a bit frustrated with the comments about reoranising the data - we have all been there. Following on from the previous posting which gave...

@ JBANKO

What part(s) of the application are you responsible for?

Are you just trying to generate reports or are you trying to improve functionality (e.g. get the application to automatically reassign...

CREATE TABLE [dbo].[JobLog](

[JobID] [int] IDENTITY(1,1) NOT NULL,

[ParentJobID] [nchar](255) NOT NULL,

[StoredProcedureName] [nchar](255) NOT NULL,

[StartDatetime] [datetime] NULL,

[EndDatetime] [datetime] NULL

) ON [PRIMARY]

ID in an identity field

No indexes

No triggers

table is currently only about 80...

Just as an aside, seeing as you are a join noobee

This will only return patients WITH telephone numbers. If you want all patients AND the telephone number if...

Can't use CTEs, we are on SQL2000 🙁

We are on SQL2000 aren't we???

Seriously though, normalising the data will make your life a lot easier. You can store it in...

@jeff Williams

The left joins are (were) only for development where I don't want to lose any records so that I can see if there are any data issues...

Taking out all the left joins and replacing with joins solves the speed, but at the risk of losing records, But I think it is acceptable losses. If the...

Thanks Phil,

I'n not sure that construct will work for me as the Coalesce() is to determine a value to be updated.

Using the Union syntax is still running down the dataset...

My final solution was to go with a single wide table with every existing combination in it. I then have a routine which checks the datasource and adds new...