Steady State killing database RRS feed

  • Question

  • Hi,

    I am running a custom .net app on an XP machine, which is logging data to an SQL 2005 PC at a very high intensity. I also have two remote PCs (running .net remoting apps) which are sending data to the main PC to be logged.

    Problem is that occasionally the main PC just locks up and stops my time critical data logging. Worse still very occasionally SQL2005 throws a wobbler and tells me that the database I am logging to has corrupt page files and must be repaired. The Db is then marked suspect.

    I have three sets of PCs working in this way, on three seperate systems. Two of them have problems, one doesn't. The differences? - The one that is unnaffected does not run steady state.

    Also, going through the event logs I find that just before any freeze or db corruption I get a message saying that the steady state service has hung.

    I have read many posts regarding hanging logons and slow boot-ups, but nothing describing what I have observed. Is is common? Should I uninstall steady state and go for some other product?


    Monday, September 29, 2008 8:43 AM


All replies


    Hi Paul, when the issue occurs, which account are you using? Administrator or standard user? Have you set any restriction on the user? You can also disable all the restrictions and then check the result and we can narrow down if the issue is Windows SteadyState related.

    Tuesday, September 30, 2008 5:06 AM
  • Hi Sean/Zhu,

    I have seen this happen with both the admin account (which is unrestricted) and the restricted user account. [Well, I was told that it happened on the admin account - I wasn't actually there. I am getting someone to attempt to remove restrictions and re-run, but this issue doesn't happen every day, so it may take a long time to get resolution one way or the other]

    The reason I suspect SteadyState is because I get this error in the event logs:

    "The Windows SteadyState Service service [sic] hung on starting"

    ... just before this error describing how my Db is knackered:

    "SQL Server detected a logical consistency-based I/O error: Incorrect checksum (expected: [foo1]; actual [foo2]). It occurred during a read of page ([foo3]) in database ID: X at offset [foo4] in file [datafile.mdf]".

    Now, I can assure you that the SteadyState service should have already started, because at the point that this fell over the .net application had been data logging for about 3 hours.

    Occasionally the PC freezes for upto two minutes at a time (which whilst running an intensive data logging process is inconvenient to say the least) and each time we can find a corresponding message in the logs saying the same thing (steadystate service has hung).

    Could the Db corruption be caused by the machine hanging..? Surely the connection would just time out?

    Thanks for your interest so far...
    Wednesday, October 1, 2008 8:14 AM

    Thank you for the update, Paul. As far as I know, Windows SteadyState or its restrictions should not affect Db on the computer and I consider the issue can be caused by other factors. I suggest you post the issue to our SQL forum via the following link:




    Thank you!

    Friday, October 3, 2008 3:42 AM
  • Thank you for your time.

    I have never seen a database corrupt like this in eight years of working on similar systems (admittedly not always as a programmer) and the only thing in common to the two systems (and unique to those two systems, too) that occasionally fail is SteadyState.

    I will certainly post on the SQL forum, in the hope that another scapegoat can be found
    Tuesday, October 7, 2008 3:14 PM