Post reply

RING_BUFFER_CONNECTIVITY Questions

  • February 11, 2018 at 4:53 am

    #329207

    I am troubleshooting some connectivity issues between our web application and SQL Server.
    When I look at ring buffer connectivity, I do see entries, but they don't always correlate to our connectivity errors in the web app.
    Here is a sample:
    https://docs.google.com/spreadsheets/d/1b7dk9AtgARAtoUuhj-pusHFCvmu9s5hezP6mf4ceVQc/edit?usp=sharing

    Since I never looked at ring buffer connectivity before our problem started, I'm not sure if we were getting these entries before the problem started.
    Can anyone render an opinion on whether or not its a big deal to see these entries when looking at ring buffer connectivity?

    Thanks

  • February 12, 2018 at 9:57 am

    #1979458

    If I were you, I will take a look at error log for it's more readable:-)

    GASQL.com - Focus on Database and Cloud

  • February 12, 2018 at 10:17 am

    #1979463

    the easiest way I've found to troubleshoot failed login attempts is to query the error log like:
    EXEC xp_readerrorlog 0, 1, N'Login Failed';

  • February 12, 2018 at 11:41 am

    #1979480

    rogsql - Sunday, February 11, 2018 4:53 AM

    I am troubleshooting some connectivity issues between our web application and SQL Server.
    When I look at ring buffer connectivity, I do see entries, but they don't always correlate to our connectivity errors in the web app.
    Here is a sample:
    https://docs.google.com/spreadsheets/d/1b7dk9AtgARAtoUuhj-pusHFCvmu9s5hezP6mf4ceVQc/edit?usp=sharing

    Since I never looked at ring buffer connectivity before our problem started, I'm not sure if we were getting these entries before the problem started.
    Can anyone render an opinion on whether or not its a big deal to see these entries when looking at ring buffer connectivity?

    Thanks

    Not necessarily with the IsSessionKilled being 1 for all of those and nothing for any of the login timers. You may want to check the record type RING_BUFFER_SECURITY_ERROR to see if there are errors with the API noted. You can find a few queries related to what you are doing in the following:
    Inside sys.dm_os_ring_buffers

    But what are the errors from the web app? That's normally where you would want to start.

    Sue

  • February 12, 2018 at 3:28 pm

    #1979539

    Thanks for the responses thus far.

    This problem may be too complicated to solve here, but I'll give it a try.
    A better ring_connectivity_log is here:
    https://docs.google.com/spreadsheets/d/1L_PwddRAyrD4qennfIgAnf15xvZdaSgOTHCDvfQOMHc/edit?usp=sharing

    The web server gets an intermittent error
    ------------------
    Error: 10054, ErrorMessage: "TCP Provider: An existing connection was forcibly closed by the remote host.
    -----------------
    This is a new problem with no changes to the client.

    The SQL server has some of the errors below which we think started about the same time, but they don't always correlate one for one with the error on the client:
    -----------------
    A fatal error occurred while reading the input stream from the network. The session will be terminated (input error: 10054, output error: 0). Error: 4014
    ------------------

    I'm going to assume that we should never see any of these errors on the SQL Server, and that they are probably related to the error message on the client.

  • February 12, 2018 at 3:56 pm

    #1979542

    rogsql - Monday, February 12, 2018 3:28 PM

    Thanks for the responses thus far.

    This problem may be too complicated to solve here, but I'll give it a try.
    A better ring_connectivity_log is here:
    https://docs.google.com/spreadsheets/d/1L_PwddRAyrD4qennfIgAnf15xvZdaSgOTHCDvfQOMHc/edit?usp=sharing

    The web server gets an intermittent error
    ------------------
    Error: 10054, ErrorMessage: "TCP Provider: An existing connection was forcibly closed by the remote host.
    -----------------
    This is a new problem with no changes to the client.

    The SQL server has some of the errors below which we think started about the same time, but they don't always correlate one for one with the error on the client:
    -----------------
    A fatal error occurred while reading the input stream from the network. The session will be terminated (input error: 10054, output error: 0). Error: 4014
    ------------------

    I'm going to assume that we should never see any of these errors on the SQL Server, and that they are probably related to the error message on the client.

    Those are network errors.  There actually is a bit of information on those as they aren't too unusual, often are related to the NIC.
    If it just started, make sure the server has been rebooted from any patches that have been applied.
    Check if updates (drivers, firmware) has been applied. Make sure the NIC drivers are up to date. Here are some of the articles:
    Troubleshooting: Connection Forcibly Closed
    TCP Chimney Setting and SQL Server Error: TCP Provider: An existing connection was forcibly closed by the remote host
    Fixing SQL Server fatal error 4014

  • February 12, 2018 at 4:00 pm

    #1979544

    There are a lot more if you search. I wasn't going to post anymore on that post as this forum software can turn everything into a hyperlink...which it did after the last one link. 🙁
    Make sure to check the OS for the SQL Server. I'll try one more post - it has more of the commands you can run to check things but make sure the NIC drivers are up to date first.
    Fixing SQL Server fatal error 4014

    Sue

Viewing 7 posts - 1 through 6 (of 6 total)

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