Spoke and Hub having problems with exchange server

Discussion in 'Cisco' started by jcharth, Oct 28, 2005.

  1. jcharth

    jcharth Guest

    Hello I setup a remote site as a spoke. For some reason they are having
    problems accessing the mailserver. OWA does not work it seems very slow
    and outlook completely freezes at the start. I was able to take over
    the remote pcs with vnc and i was able to browse remote server shares.
    Any ideas what the problem could be? thanks
     
    jcharth, Oct 28, 2005
    #1
    1. Advertisements

  2. jcharth

    Merv Guest

    You need to provide a LOT more info befor anyone will be able to help
    you.

    - router models at each hub and spoke
    - IOS versions on both routers
    - WAN transport and bandwidth
    - routing protocol used on link
    - show interface from both routers for the link between the two sites
    - show interface switching
    - show ip cef
    - etc, etc
     
    Merv, Oct 28, 2005
    #2
    1. Advertisements

  3. jcharth

    nazgulero Guest

    Hello,

    it could just be that the line between your remote site and your main
    site is saturated...what happens if you let the clients select the
    'Basic' option on their login screen ? Can you check if the client's
    mailboxes are very large, or create a new client and have that client
    try to access OWA from the remote site ?

    Regards,

    GP
     
    nazgulero, Oct 28, 2005
    #3
  4. jcharth

    jcharth Guest

    Thanks
    The router models are as follow
    ios 12.3 1700
    ios 12.1 1700
    pix 515 6.3

    i can browse the print server website, but even owa seems to stall. May
    be i should remove my traffic shapping, just for the day.
     
    jcharth, Oct 28, 2005
    #4
  5. jcharth

    jcharth Guest

    Hello I did a mtr to a remote host, if i increase the ping size to 1500
    or 1250 i get 10 hups, if i decrease it to 1000 i get 3 hubs, any idea
    how to fix this?
    thanks.
     
    jcharth, Oct 29, 2005
    #5
  6. jcharth

    nazgulero Guest

    Hello,

    weid indeed...if the number of hops depends on the size of the packet,
    that must mean that somehow the intermediate devices route traffic
    based on packet size. Can you post the MTR results with both packet
    sizes, if possible ? I am interested in the intermediate hops...by the
    way, which country are you in ?

    Regards,

    Naz
     
    nazgulero, Oct 30, 2005
    #6
  7. jcharth

    jcharth Guest

    with packet of 1492 i get
    10.10.2.1 1 0% 8 8
    No response 2 100% 7 0
    10.10.11.1 3 58% 7 3
    No response 4 0% 1 0
    No response 5 0% 1 0
    No response 6 0% 1 0
    10.10.11.1 7 0% 1 1


    with packet of 1000
    i get

    10.10.2.1 0% 16 16
    No response from host 100% 15 0
    10.10.11.1 7% 15 14
    10.10.11.1 0% 1 1


    The columns are lost% sent and received. I wish i could do the test
    with a host inside the 10.10.11.x subnet but there are no hosts online
    today.
    The results look similar but the last row is 10.10.11.47 instead. I
    tried changing the MTU on the interface inside and on the clients but
    that did not work, i think i set it up too low to 576, I am thinking
    may be i can change the mtu on the mailserver or just increase it to
    1000 on the internal interface of the router. Thanks.
     
    jcharth, Oct 30, 2005
    #7
  8. jcharth

    jcharth Guest

    I havent test this but someone just mention to put this in the router's
    internal iterface
    ip tcp adjust-mss 1380
     
    jcharth, Oct 31, 2005
    #8
  9. jcharth

    jcharth Guest

    changing the mss did not work, i had to change mtu and mss it seems
    like that does the trick.
     
    jcharth, Nov 3, 2005
    #9
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.