Managing / Administering 200+ SQL servers

  • You are in a tough spot but my suggestion is to step back and do some cost analysis. Find out the total cost of the infrastructure they are going to put in through 200 sites and then compare that versus the cost analysis of a more consolidated approach. Part of the reason for getting more consolidated is to drive prices down. So if they think it is cheaper to have more servers you need to prove wether or not they are working with the right assumptions. Also in that kind of environment you are going to be doing alot of data moving. That needs to be worked into cost and will probably require additional resources to support it. Needless to say you are probably out of compliance somewhere and once Microsoft comes knocking to pay the full bill you can be the cool head pointing to how to solve the issue.

  • As I was told by Domain Admins:

    "You should accept current situation, in the near future infrastructure will not be changed"

    also they told me that long before me question about Server consolidation was risen, but with no good ending.

    Boses acceps prices for the licenses and don't wish to discus about changing something in structure.

    I gues I'll have to accept being more then 200 SQL servers DBA.

  • I would make them in writing define the role of DBA for you. You might be in a lose/lose situation and you don't even know it. If you are the DBA, then while you may not be able to dictate somethings like consolidating, you should be able to reorganize things to make your life easier. Some example things I would consider is getting to a standardized installation on all 200 machines. This is a big task, but it pays huge dividends. You can log into any of my servers, and they are all identical. You never have to wonder where my logs are, where my datafiles are, where the binaries are installed. This helps in enterprise wide automation. I can write a vb task to scan over a list of servers in a text file and perform operations against them using WMI or other methods since they are basically cookie cutter machines.

    Jonathan Kehayias | Principal Consultant | MCM: SQL Server 2008
    My Blog | Twitter | MVP Profile
    Training | Consulting | Become a SQLskills Insider
    Troubleshooting SQL Server: A Guide for Accidental DBAs[/url]

  • As far as I know, the servers are divided to three groups, according to the shop size,

    in these groups servers have identical Hardware + Software.

    so there are already some Standards.

    p.s.

    still trying to figure out whats good / bad in Ms Operations Manager.

    there are lots of things to look at, I'd say too much 🙂

  • For 1$ I could give you an advice..set up jobs that alert you with disk space or any errors reporting in the errorlog.

    We use an application like 'appmonitor' that queries all our DB and app servers and returns notifications via email. Little bit of work, but you would not have to pay anything. Also the manual process is cumbersome, and would become a nightmare when you actually get all those 200 servers up and running..

    Thanks!!

    The_SQL_DBA
    MCTS

    "Quality is never an accident; it is always the result of high intention, sincere effort, intelligent direction and skillful execution; it represents the wise choice of many alternatives."

  • You know something I thought about reading the last post. At our local user group one of the guys presented on monitoring SQL Server with RSS. He used a product call RSS Reporter from xSQLSoftware to monitor hundreds of servers:

    http://www.xsqlsoftware.com/Product/Sql_Server_Rss_Reporter.aspx

    something to look at and consider, as the licensing isn't to bad for it.

    Jonathan Kehayias | Principal Consultant | MCM: SQL Server 2008
    My Blog | Twitter | MVP Profile
    Training | Consulting | Become a SQLskills Insider
    Troubleshooting SQL Server: A Guide for Accidental DBAs[/url]

  • One more add-on:

    Try to set up the requirements for database support on your side, including server version/patch level, user access, SLA agreement (if any), etc. This could be a short or long list.

    1. Support those meet the requirements;

    2. Share responsibilities for those partial meet the requirements;

    3. Do not support those (if applicable in your shop) that have not met the critical criteria (SQL2000 SP3 under, everyone has SA privileges/OS admin, no business owner, etc).

  • In addition to what has been mentioned, I would definitely try to make sure that you can have the sql servers mail you. And if jobs are failing on one server then make sure there is something distinctive about that servers emails to make it easier to find. As for SP level, you will need to get up to SQL 2000 SP 4 to get support from microsoft if something goes really wrong.

    Good Luck

Viewing 8 posts - 16 through 22 (of 22 total)

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