SQL Server Performance

Windows Task MangerTab Networking - Network Utilization jumps to 97%

Discussion in 'Getting Started' started by lcerni, Sep 8, 2009.

  1. lcerni New Member

    Looking at Windows Task MangerTab NetworkingAbout every 7 minutes the Network Utilization jumps to 97%. I am trying to figure out why. However, I feel like I am looking for a needle in a haystack. I have tried performance monitor and SQL Server Profiler. Perhaps I am using the wrong tools or the wrong counters. Can anyone provide assistance? Which tool and counters should I be looking at?
  2. moh_hassan20 New Member

    This is the traffic passing through the network card, which may be due to any running process on the server

    if you suspect sql server is the reason , you can monitor the following network counters using performance monitor tool:
    • Server: Bytes Received The number of bytes the server received from the network. This counter is the indicator of how busy the server is.
    • Server: Bytes Transmitted The number of bytes the server has sent to the network.This counter is of how busy the server is.
    • Network Inter: Output Queue Length The length of the output packet queue (in packets).It should not be longer than 2, otherwise delays are being experienced and there is a network bottleneck which should be resolved.
  3. lcerni New Member

    For Server: Bytes Received and Server: Bytes Transmitted I got zero for the values.
    For the Network Interface I got the following values:
    Monday
    20090914 11:25am - 20090914 12:24pm
    ObjectCounterInstance AvgMinMax
    Network InterfaceOutput Queue LengthBroadcom BCM5708C NetXtreme II GigE [NDIS VBD Client] 0029
    Network InterfaceOutput Queue LengthMS TCP Loopback interface 000
    Network InterfaceBytes Received/secBroadcom BCM5708C NetXtreme II GigE [NDIS VBD Client] 199666.9459597.5931149878
    Network InterfaceBytes Received/secMS TCP Loopback interface 109.91102556.094
    Network InterfaceByptes Sent/secBroadcom BCM5708C NetXtreme II GigE [NDIS VBD Client] 411202512813.45698884204
    Network InterfaceByptes Sent/secMS TCP Loopback interface 109.91102556.094
    Network InterfaceCurrent BandwidthBroadcom BCM5708C NetXtreme II GigE [NDIS VBD Client] 100000000010000000001000000000
    Network InterfaceCurrent BandwidthMS TCP Loopback interface 100000001000000010000000
    Network InterfaceBytes Total/secBroadcom BCM5708C NetXtreme II GigE [NDIS VBD Client] 431200925036.55899540684
    Network InterfaceBytes Total/secMS TCP Loopback interface 219.82205112.188I am not sure what these numbers mean. Does this mean I have a SQL Server issue or does this mean I have an issue with the Network Card? Should I increase the 'network packet size (B)' from 8192 to something higher?
  4. moh_hassan20 New Member

    your network is healthy and there is no network bottleneck because:
    Output Queue Length for sent/receive =0 (good indicator)

    your interface is 1 Gbs, and Bytes Total/sec = received + transmitted = 4312009 byte = 4.3 M (not busy at all)
    There is no need to change the default packet 8192K , keep it.

  5. lcerni New Member

    Then does that mean that when the Network Utilization jumps to 97% every 7 minutes that this is normal for this server and that I shouldn't worry about it?
  6. satya Moderator

    Do you have any network monitoring tools or alert (software) installed on this machine?
  7. lcerni New Member

    No, we do not have any network monitoring tools or alert (software) installed on this machine. I have noticed that under the Network Adaptor History it is the Bytes Sent (Red) that is spiking. My manager is going to ask the LAN Team to use their tools to follow up on this. He feels that I have gone as far as I can go and that their tools can "sniff" out what is going on.
  8. satya Moderator

    I believe you should rely on PERFMON tool to get optimum values than going with Task Manager on the server. Also monitor the number of users and db usage regularly.

Share This Page