back to article Violin, Microsoft slip Windows, SQL Server, apps, flash into box

Violin Memory has turned its converged server-storage 6000-series flash array into an engine for SQL Server which is capable of putting data directly into an app server's memory – thanks to a little help from Microsoft. The 6000-series all-flash array runs Windows Storage Server 2012 R2 on two integral blade servers. It …

COMMENTS

This topic is closed for new posts.
  1. Anonymous Coward
    Anonymous Coward

    Interesting to see what sort of Benchmarks this puts out.

    SQL Server is already wiping the floor with the competition in terms of performance:

    http://www.tpc.org/tpch/results/tpch_perf_results.asp

    http://www.tpc.org/tpce/results/tpce_perf_results.asp

    http://www.tpc.org/tpcvms/default.asp

    1. h4rm0ny

      I honestly could not say if MS SQL Server performance is better or worse than its competitors, but I feel confident in saying that the differences are not so great that any could be said to be "wiping the floor" with the other. For a start, fair comparisons between different databases are pretty hard to set up.

      I think Robert Haas who is chief architect for Postgres put it best in a blog post response to someone asking him about SQL Server vs. Postgres performance. I've dug it out for you. (Caveat: I am a major postgres fan).

      http://blogs.enterprisedb.com/2010/12/20/postgresql-performance-vs-microsoft-sql-server/

      It's only short and worth reading (imo), but for those who can't be bothered here is a paragraph from the end:

      I think the important thing to realize here is that both PostgreSQL and Microsoft SQL Server are complex, robust products with many features and generally good performance. People can and do use both products to manage large amounts of data in critical production environments. It would be a serious mistake to believe any categorical statement about which system is faster and by how much. There are going to be cases – in either system – where things just don’t perform well.

      He is a very, very smart person.

      1. h4rm0ny

        Okay, seriously - whoever is following me from story to story and downvoting everything I post, have you any idea how sad that is? I mean look at you - you're disagreeing with a post saying that comparing different database types is hard. Why should that offend you? Because it certainly isn't wrong.

      2. Anonymous Coward
        Anonymous Coward

        You are dreaming if you think Postgre SQL is even remotely comparable with MS SQL Server. It's like comparing Open Office with Office 2013. It might do if you need the bare essentials, but it's in no way fit for most enterprises.

        And that worthless quote is from nearly 4 years ago.

  2. Anonymous Coward
    Anonymous Coward

    Not sure if you're being sarcastic with this comment, or serious. Don't get me wrong, I'm a MS SQL Server DBA and developer, but stating MS SQL Server is "wiping the floor with the competition" and then referencing two benchmarks that have only MS SQL Server benchmarks and nothing else to compare them to (TPC-E and TCP-VMS), and another where MS SQL Server is top in 3 out of 6 benchmarks, is hardly what I'd call statistically significant data to base that conclusion on ...

    1. Anonymous Coward
      Anonymous Coward

      You can't be a very well informed SQL DBA / Developer then or you would understand the significance of those benchmarks.

      Here it is spelt out in crayon for you:

      http://www.onwindows.com/Article/hp-sets-world-data-warehousing-records-on-microsoft-sql-server-2014-17044#.U1bA8HnQe00

      Better performance at half the cost. And with ~ an of magnitude fewer security vulnerabilities to worry about too on Windows / SQL compared to Solaris / Oracle DB.

      1. Anonymous Coward
        Anonymous Coward

        ^Should say '~ an order of magnitude fewer'

      2. Anonymous Coward
        Anonymous Coward

        Next time try reading before posting. I'm not disputing that SQL Server holds records, or that's it has high performance vs cost, what I'm disputing is refering to links with no comparison data as showing how good SQL Server is.

        Go to the post I was replying to, and follow the links provided. You can confirm for yourself that in two of them there's no comparison with a non-SQL Server product (all the benchmarks are for various versions of SQL Server), and that the other has 3 out of 6 benchmarks held by non-SQL Server products.

        Shame to see that reading comprehension doesn't factor into your daily routine ...

  3. Anonymous Coward
    Anonymous Coward

    Sometimes news is not all that new

    http://www.theregister.co.uk/2013/10/16/violin_memory_enters_the_server_business/

  4. Stephen Channell
    Happy

    Topping the 10Tb TPCH with commodity SQL/Server & a DL580 is spanking proporations

    Not very long ago TPC-H was dominated by custom column-stores databases that were almost impossible to get into a production data-centre.. those column store databases had surpassed the astronomically expensive Teradata MPP kit that also had trouble competing with Oracle & DB2, despite clear performance advantages.

    That a commodity HP DL580 box that's on most big shops "standard server list" can beat custom kit, supported by legions of SQL/Server DBAs (that you probably already pay) is spanking proportions... if the Violin boxes with RDMA right into SQL/Server bufferpools work like they say, we should see some really incredible TPC performance numbers that should spank the "fake" Oracle TPC-C scores (real OLTP performance is not measured in the number of brocade switches)

    1. Anonymous Coward
      Anonymous Coward

      Re: Topping the 10Tb TPCH with commodity SQL/Server & a DL580 is spanking proporations

      It's a DL980 actually, but doesn't really change your point...

This topic is closed for new posts.