May 9, 2017 at 12:00 am
Comments posted to this topic are about the item Best Option for Working with Files in SSIS
May 9, 2017 at 8:23 am
Thanks for sharing the results of this experiment.
I'd be extremely interested in discovering WHY things slowed down when they were ran in parallel.
Is it possible that the storage system was simply overwhelmed? Disk seeks were hammering your disk because files were in two different places?
May 14, 2017 at 4:19 am
Ahhh... don't your results and conclusion contradict themselves?
Results:
"The File System task is the worst option and the slowest one with 130 seconds"
Conclusion:
"As you can see, the fastest solution is the File System Task."
May 14, 2017 at 1:32 pm
Brent Leslie - Sunday, May 14, 2017 4:19 AMAhhh... don't your results and conclusion contradict themselves?Results:
"The File System task is the worst option and the slowest one with 130 seconds"Conclusion:
"As you can see, the fastest solution is the File System Task."
It IS a bit confusing until you realize that he was speaking of the non-parallel versions in his conclusion. It would have been handy if he pointed that out in that statement but he did clarify a bit further on in the Conclusion.
When we talked about the tasks in parallel, we did not get a better performance except with the cmd. In few words, in this case, having the process in parallel does not guarantee a better performance
--Jeff Moden
Change is inevitable... Change for the better is not.
May 14, 2017 at 3:00 pm
Except for the previously stated bit of confusion in the Conclusion of the article, I liked the article. It was simple, straight forward, and the experiment appears to be repeatable for any one that takes the time to try it. Thanks for taking the time to post it, Daniel.
--Jeff Moden
Change is inevitable... Change for the better is not.
November 22, 2019 at 2:33 pm
I'm curious what the impact of reading/writing to the same disk had on the parallel execution. Would results improve by writing to another physical disk?
November 22, 2019 at 9:01 pm
If there are multiple files, an analysis of ROBOCOPY would be interesting with various /MT settings.
November 22, 2019 at 11:32 pm
Might also want to consider BCP and or bulk insert, sometimes it's really fast on imports..... have to see whats faster in situations with bigger files.
December 31, 2021 at 8:55 am
This was removed by the editor as SPAM
December 31, 2021 at 11:23 am
MicrosoftSQL Server Integration Services (SSIS) is a platform for building high-performance data integration solutions, including extraction, transformation, and load (ETL) packages for data warehousing.
What's your point and how does it relate to the article?
The absence of evidence is not evidence of absence
- Martin Rees
The absence of consumable DDL, sample data and desired results is, however, evidence of the absence of my response
- Phil Parkin
January 1, 2022 at 5:54 pm
It's a direct quote from here...
Must be a spammer setting up.
--Jeff Moden
Change is inevitable... Change for the better is not.
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply