When creating a DTS package that copies multiple tables from one sql
server to another, is there an advantage to copying tables
simultaneously as opposed to using on-completion steps and copying the
tables one at a time? I would think that the server could copy one
table at a time just as fast as it could copying multiple tables at
the same time, as the amount of data moved is the same using either
method. I read somewhere that the server may be faster doing the
tables asynchronously. Is this true?

Walter

In article <28db1398.0309020544.746d19b8 (AT) posting (DOT) google.com>, Walter
Mallon <waltmallon (AT) yahoo (DOT) com> writes

It depends. It should be faster in parallel, but it may make no
difference, or maybe even slower due to contention "arguments".

If you go for parallel running tasks without the "On Completion"
constraint they still may not run in parallel. The biggest issues are
disk IO, network and memory, as they normally become saturated well
parallel processing becomes an option.

There are other issues such as a Single SQL Server OLDE-DB connection
will serialise all work given to it, so use multiple connections.

If you have specific issues can you expand.

--
Darren Green (SQL Server MVP)
DTS - http://www.sqldts.com

Thanks for the response. I don't have a specific issue. I was
developing a large DTS package and was creating it so that one table at
a time was copied. I've seen some DTS packages that don't use the
completion steps and just have a number of transform data tasks
executing together. I was just curious if there was a definate
difference and if one method was preferred over the other.

Thanks again,
Walter


*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

In article <umYIdljcDHA.2508 (AT) tk2msftngp13 (DOT) phx.gbl>, Walter Mallon
<waltmallon (AT) yahoo (DOT) com> writes
I think there is no correct answer. Use what is easier for you to
create, or use what is easier for you to manage or use what is best for
performance, depending on your priorities in this instance. For
performance, you will need to test your system as there are so many
variables it is impossible for me or anyone else to say which method
will work best for you.

--
Darren Green (SQL Server MVP)
DTS - http://www.sqldts.com