Data Compression Objects

  • Comments posted to this topic are about the item Data Compression Objects

  • Nice solid yet simple enough question, thanks

    Hope this helps...

    Ford Fairlane
    Rock and Roll Detective

  • I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    Need an answer? No, you need a question
    My blog at https://sqlkover.com.
    MCSE Business Intelligence - Microsoft Data Platform MVP

  • Koen Verbeeck (12/17/2013)


    I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    Whats your performance been like since ?

    Hope this helps...

    Ford Fairlane
    Rock and Roll Detective

  • Nice question, thanks.

    Igor Micev,My blog: www.igormicev.com

  • +1 Steve. Thanks

  • Ford Fairlane (12/17/2013)


    Koen Verbeeck (12/17/2013)


    I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    Whats your performance been like since ?

    I was about to ask this same thing. I have been considering doing this as well. So far I have only added page compression to our archive tables and not to ones being accessed more frequently.

  • Koen Verbeeck (12/17/2013)


    I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    What kind of compression ratio have you obtained?

    [font="Verdana"]Please don't go. The drones need you. They look up to you.[/font]
    Connect to me on LinkedIn

  • KWymore (12/17/2013)


    Ford Fairlane (12/17/2013)


    Koen Verbeeck (12/17/2013)


    I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    Whats your performance been like since ?

    I was about to ask this same thing. I have been considering doing this as well. So far I have only added page compression to our archive tables and not to ones being accessed more frequently.

    If the compression brings you over 70% storage gain, than you should compress without any measurements, otherwise you should make some measurements concerning the CPU usage on compressed and uncompressed tables.

    Once I did some investigation on compression on a small system. I compressed the entire database (mixed: page and row) and run a tool that was processing data against compressed and uncompressed databases. The processing time was same, and i obtained a storage gain. The avg CPU usage increased but not for much. It's recommended the avg CPU should not be more than 80%. So you can even compress the more frequently accessed tables.

    If the reads operations dominate over the writes (e.g. more than 95:5) than you go with compression.

    Regards

    IgorMi

    Igor Micev,My blog: www.igormicev.com

  • IgorMi (12/17/2013)


    KWymore (12/17/2013)


    Ford Fairlane (12/17/2013)


    Koen Verbeeck (12/17/2013)


    I have been enabling page compression on our entire data warehouse recently, so this was a pretty easy question 🙂

    Whats your performance been like since ?

    I was about to ask this same thing. I have been considering doing this as well. So far I have only added page compression to our archive tables and not to ones being accessed more frequently.

    If the compression brings you over 70% storage gain, than you should compress without any measurements, otherwise you should make some measurements concerning the CPU usage on compressed and uncompressed tables.

    Once I did some investigation on compression on a small system. I compressed the entire database (mixed: page and row) and run a tool that was processing data against compressed and uncompressed databases. The processing time was same, and i obtained a storage gain. The avg CPU usage increased but not for much. It's recommended the avg CPU should not be more than 80%. So you can even compress the more frequently accessed tables.

    If the reads operations dominate over the writes (e.g. more than 95:5) than you go with compression.

    Regards

    IgorMi

    Thank you, I will keep that in mind when I am able to get back to that testing!

  • Our data warehouse got a compression ratio of 68% using page compression. The best compression ratios are usually with star schemas with all (or close to all) integer surrogate keys. However, beware of compression's effect on data loading time. If you already are squeezing your data load into a small window of time or if your data load is CPU bound, you MAY be in for a surprise.

    See The Data Loading Performance Guide for how compression affects CPU usage/data loading time and much, much more. It was created for SQL Server 2008, but it remains relevant today.

    http://technet.microsoft.com/en-us/library/dd425070(v=sql.100).aspx

  • Elementary dear Watson, er, Jones. 🙂

    Thanks, Steve!

  • This was removed by the editor as SPAM

  • Nice tidy question on a well documented issue, so I'm surprised that more than 30% so far have it wrong.

    Tom

  • Nice question....

Viewing 15 posts - 1 through 15 (of 18 total)

You must be logged in to reply to this topic. Login to reply