none
SNA Warning 14 and 15 on HIS 2010 SP1 RRS feed

  • Question

  • Greetings,

    while conducting the SNA server stress test, we found two warning message was logged in the windows event log:-

    1. First Warning Message

    Event ID: 14 (Critical Buffer Congestion)
    Source: SNA IP-DLC Link Service
    Cause: The Buffer Manager Monitor detected critical buffer congestion. This condition can occur if the node receives a large, sudden burst of data traffic on several sessions.
    Effect: The IP-DLC link service will throttle back session traffic by withholding pacing responses or resetting pacing windows on adaptive-paced sessions until the congestion has reduced.
    Action: The condition should clear itself automatically, but if it occurs frequently, it may indicate that the node is routing more sessions than it can handle efficiently with the available buffer storage.  Investigate system load.

    2. Second Warning Message

    Event ID: 15 (Buffer Congestion Relieved)
    Source: SNA IP-DLC Link Service
    Cause: Buffer congestion relieved.
    Effect: The IP-DLC link service will stop throttling back session traffic.  If the congestion eases further then the IP-DLC link service will start to increase pacing windows on adaptively paced sessions.
    Action: If buffer congestion occurs frequently, it may indicate that the node is routing more sessions than it can handle efficiently with the available buffer storage.  Investigate system load.


    Any possible way to reproduce this symptoms so that we can write the code to handle this scenario?

    Any feedback is greatly appreciated.

    Thank You

    Regards,

    Viknes


    Viknes Raj

    Wednesday, April 15, 2015 10:17 AM

Answers

  • The only way that I know of to reproduce these symptoms is to send high volumes of traffic to the HIS Server to attempt to fill up the buffers that handle the incoming data. Overall, the IP-DLC link service and IBM Enterprise Extender on the mainframe side should be able to deal with these situations when they occur. The underlying HPR protocol is designed to throttle traffic based on an internal algorithm to deal with high traffic, slow network type issues.

    Thanks...


    Stephen Jackson - MSFT

    Tuesday, April 21, 2015 4:47 PM