locked
CRM outlook client doing horrible things to network bandwidth and latency RRS feed

  • Question

  • Thanks in advance to anyone who can help me on the trail to solving this problem, I haven't been able to google up any answers on this one.  We've been using crm online for over a year, and have tried a couple of times to get the outlook client going with minimul success.

    We currently have a small office with a couple servers, 3 office workers, and 2 more that connect from outside over a VPN.  We have a pretty standard AT&T '6GB' DSL connection.  A normal baseline speed test on the line yields 32ms latency, 600-700 Kb/s upload speed, and 3,000-5,000 KB/s download speed.

    We have 1 user that is using the CRM Outlook client on an IBM Laptop running Office 2010, Windows 7.  CRM client wias installed with the default options. Our CRM online implementation has a handfull of custom data fields, but is out-of-the-box otherwise. Usage of CRM online over the web interface does not seem overly slow. IBM Laptop has all current service packs.

    The issue is that about 3 minutes or so after he starts outlook, another network speed test yields 400-1200ms latency, 284 Kb/s upload speed, and 178 KB/s download speed.  This impacts the network so badly that it pretty much is unusable for all of the other users in the office.  As long as his outlook client is up and running, the network continues to perform extremely badly.

    Here is a list of what I've tried to resolve the issue:

    Problem originally occurred on Windows Vista and Office 2007. Performed a fresh install of windows 7 (and office 2007), but problem still occurred. I just upgraded it to Office 2010, but problem the still occurs. We disabled all outlook plugins except for CRM. Disabling the CRM plugin makes the problem go away. We tried shutting down all non-essential windows services, plugins, and startup programs. Very consistently, starting the CRM outlook client makes the problem happen, and closing outlook or disabling the CRM for outlook plugin makes it go away.

    We have thought it may be an initial synchronization for the offline client, but have let it run constantly over several weekends, and it's currently been running approximately 10 hours per day for at least 3 days at this point.  Our CRM implementation is fairly small at 0.86GB, so I hardly think volume is an issue. I run a ping -t to an IP on the internet to monitor the latency pretty much all day, and the latency is pretty consistently 300-900ms as long as the outlook client is running.  There seems to be no pattern related to time or frequency with the latency, it's pretty consistent all day long.

    Anyone have any ideas?

     

    • Moved by Xiaopeng Wu Tuesday, March 8, 2011 10:10 PM Forum consolidation (From:CRM Online)
    Tuesday, October 5, 2010 5:58 PM

Answers

All replies

  • First, make sure you're running the latest version of the client.  You can get the latest version here: http://www.microsoft.com/downloads/en/details.aspx?FamilyID=e8e6c81a-b7b2-4e57-8599-604967617396&displaylang=en

    If that doesn't fix the problem, please open a support ticket.  That's definitely not expected behavior, and one of our support engineers can help you debug it.  https://mbs.microsoft.com/support/createincident.aspx


    --Matt Cooper, Microsoft Dynamics CRM
    Wednesday, October 6, 2010 2:12 AM
    Moderator
  • Hi, did you get your problem solved? We are experiencing similar issues with the 2011 version.

    Krzysztof
    Thursday, August 11, 2011 11:58 AM
  • The problem is much less severe with the 2011 version, however, it's still continues to plague us.  We have narrowed the problem down to be related to the CRM contact management, where e-mail in a users inbox is scanned to attach to CRM records. We use both exchange online and crm online, so it seems that the CRM process makes a request to the external exchange online server, which doubles up the outside network traffic. I think if we were using an internal exchange server with the CRM connector configured, the problem would likely not be so bad.
    Thursday, August 11, 2011 1:24 PM
  • Thanks for the info. In our case so far the excess traffic is attributed to background synchronization that occurs between the client and server, in particular all traffic to and from organization.svc. We are having onsite installations of both CRM and Exchange tough.

     

    Thursday, August 11, 2011 1:40 PM
  • HI all,

     

    We are in the process of implementing MS Dynamics 2011, we are experiencing the same issue. Like Kyrysztof we have onsite implementations of Exchange (2007) and Dynamics 2011. We are running windows 7 and office 2010, SQL 2008 clustered with VMWARE front end serves. We have only just discovered this issue when performing offsite testing via our WAN. In 1 hour my client generated 350MB of traffic (I have installed Fiddler http) most of which is associated to / australia/xrmservices/2011/organisation.svc. This traffic is generated continually even when you are not touchning the machine and dont have dynamics showing via outlook.

    We are not seeing this issue via the web client. Steps we have taken

    Run up physical box, validated bottle necks in LAN / WAN / VMware vitural nics etc, no issues.

    Updated QoS on WAN

    Run up stock standard organisation with no customisations, still happens

    Disabled SSL so that fiddler could intersect.

    Disabled replication / address book  via CRM option screen

    Determined that it relates to XML / Soap but cant read as this still seems encrypted

    # Result Protocol Host URL Body Caching Content-Type Process Comments Custom
    45 200 HTTP   /Australia/XRMServices/2011/Organization.svc 3,064 private application/soap+xml; charset=utf-8 outlook:952    

    body size varies from 3064 bytes to 8917544 bytes

    See details below from one of the http requests whilst on the LAN

     

    Request Count: 1

    Bytes Sent: 6,614 (headers:217; body:6397)

    Bytes Received: 8,917,779 (headers:235; body:8917544)

    ACTUAL PERFORMANCE

    --------------

    ClientConnected: 14:51:40.741

    ClientBeginRequest: 14:51:40.772

    ClientDoneRequest: 14:51:41.131

    Determine Gateway: 0ms

    DNS Lookup: 0ms

    TCP/IP Connect: 0ms

    HTTPS Handshake: 0ms

    ServerConnected: 14:51:37.933

    FiddlerBeginRequest: 14:51:41.131

    ServerGotRequest: 14:51:41.131

    ServerBeginResponse: 14:51:41.131

    ServerDoneResponse: 14:51:53.738

    ClientBeginResponse: 14:51:53.754

    ClientDoneResponse: 14:51:53.785

    Overall Elapsed: 00:00:13.0129830

    RESPONSE CODES

    --------------

    HTTP/200: 1

    RESPONSE BYTES (by Content-Type)

    --------------

    application/soap+xml: 8,917,544

    ~headers~: 235

    Any help would be appreciated, I have a meeting with Microsoft on the 17th november and will give an update. Currently we dont have a hope in hell deploying this to remote sites over the WAN

     Had the meeting with MS, the issue relates to pinning views, this is currently with the development team at microsoft. Microsoft feel that this is by design. Will update when i have their feedback.

    Monday, November 14, 2011 4:25 AM