locked
High RPC Average Latency (backend) RRS feed

  • Question

  • I am currently in the process of migrating from an Exchange 2007 to 2010 environment and use Quest to monitor the infrastructure.
    I am receiving a number of alerts relating to MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend).
    The performance counters are running against my 2010 CAS servers and consistently producing a value higher than my threshold of 250.
    The alert appears to be related to a CAS server in one site having latency with a mailbox server in another site.
    I was wondering if anyone else has seen this issue and if this is normal.

    Example of the type of alert is as follows:


    ServerAclientaccess

     
     

    Performance Counters

       

    Name:

    Threshold:

    New Value:

    Previous Value:

       

    MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend) (ServerBmailbox.remotesite.com)

    250 

    1264.00

    11282.00282.00

       

    MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend) (ServerCmailbox.remotesite.com)

    250 

    8959.00

    8959.00



    • Edited by MJH74 Friday, October 7, 2011 9:56 PM
    Friday, October 7, 2011 9:54 PM

All replies

  • First question to ask is there there is a CAS in one place talking to an MBX in another place.
    I shall ignore the semantics of the word “site” in the context of Active Directory for this first question.
    If you CAS to MBX in the location where the CAS is local to the MBX what do you get?
     
    We can’t fix your problem until you can clarify to us why your words are suggesting that you are outside of a supported configuration. Once understood, we can move forward.
     
     
     
    "MJH74" wrote in message news:8adc8613-d21c-42e3-a66c-3c7dd1f0e80e...

    I am currently in the process of migrating from an Exchange 2007 to 2010 environment and use Quest to monitor the infrastructure.
    I am receiving a number of alerts relating to MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend).
    The performance counters are running against my 2010 CAS servers and consistently producing a value higher than my threshold of 250.
    The alert appears to be related to a CAS server in one site having latency with a mailbox server in another site.
    I was wondering if anyone else has seen this issue and if this is normal.

    Example of the type of alert is as follows:


    ServerAclientaccess

     
     

    Performance Counters

       

    Name:

    Threshold:

    New Value:

    Previous Value:

       

    MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend) (ServerBmailbox.remotesite.com)

    250 

    1264.00

    11282.00282.00

       

    MSExchange RpcClientAccess Per Server\RPC Average Latency (Backend) (ServerCmailbox.remotesite.com)

    250 

    8959.00

    8959.00




    Mark Arnold, Exchange MVP.
    Sunday, October 9, 2011 12:03 AM
  • Yes, I have a CAS in one place talking with a MBX in another place.  Which I find odd as in each site (direct meaning to AD) I have a CAS and MBX local to that site.  Furthermore my question would be, why does a CAS in one site need to speak with a MBX in another site?

    The CAS to MBX in the location where the CAS is local to the MBX I have no latency.

    Monday, October 10, 2011 2:37 PM
  • That’s fine. You need a CAS role on the MBX or on it’s own box next to the MBX on the other site. I would add the CAS role to the other box and then do the research on CAS Redirection and Proxy. If you have users in that site you would without a doubt need to have that CAS so you are not achieving anything by not having that role on or near the remote MBX.
     
     
    "MJH74" wrote in message news:08614918-d85f-4071-88eb-bca8f0a7d27b...

    Yes, I have a CAS in one place talking with a MBX in another place.  Which I find odd as in each site (direct meaning to AD) I have a CAS and MBX local to that site.  Furthermore my question would be, why does a CAS in one site need to speak with a MBX in another site?

    The CAS to MBX in the location where the CAS is local to the MBX I have no latency.


    Mark Arnold, Exchange MVP.
    Monday, October 10, 2011 6:17 PM
  • After re-reading my last post I believe I need to clarify my setup.

    Every site I have a MBX role, I have an associated CAS role in the same site.

    My previous comment of, "Yes, I have a CAS in one place talking with a MBX in another place." means that there is a MBX in one site contacting a CAS in another site, even though there is a CAS in the same site as the MBX.

    I am wondering why there is contact between a MBX in one site to a CAS in another site, even though there is a CAS in the same site as the MBX.

    Hope that helps clarify what I'm looking for.

    Wednesday, October 12, 2011 7:04 PM
  • You would want to make sure that your AD Sites and Services are correctly set up and that the subnets are all set properly.
    Then see what it is that is working slowly. You can obviously see the latency on the box but you probably need to drill down and try isolate what that RPC traffic is in response to. Is it user mailboxes. Is it PF traffic, or something else entirely.
     
     
    "MJH74" wrote in message news:95e5ec9a-a6bb-47fe-a54d-1a61a772dfb0...

    After re-reading my last post I believe I need to clarify my setup.

    Every site I have a MBX role, I have an associated CAS role in the same site.

    My previous comment of, "Yes, I have a CAS in one place talking with a MBX in another place." means that there is a MBX in one site contacting a CAS in another site, even though there is a CAS in the same site as the MBX.

    I am wondering why there is contact between a MBX in one site to a CAS in another site, even though there is a CAS in the same site as the MBX.

    Hope that helps clarify what I'm looking for.


    Mark Arnold, Exchange MVP.
    Saturday, October 15, 2011 5:02 PM
  • MJH74,

    I'm having the same problem with the alert due to the CAS servers trying to connect to a remote MB server.  The remote MB server has their own CAS, so I'm not quite sure why I keep getting this  How did you resolve your issue?

    Monday, March 11, 2013 10:19 PM