Re: Strange networking issue

From: Bruce Zimmer (b.r.zimmer_at_WORLDNET.ATT.NET)
Date: 04/25/03

  • Next message: Klaus Oberle: "Re: hdisks missing"
    Date:         Fri, 25 Apr 2003 09:06:52 -0400
    To: aix-l@Princeton.EDU
    
    

    Yes same subnet, but what subnet is the 10/100 interface on. If they
    are both on 170.29.112/24 then the packet will go from 170.29.112.254 to
    170.29.112.30, but the return packet will exit the AIX machine from
    170.29.112.xxx (the 10/100 interface) and if the 10/100 and gig are not
    on the same physical network the packet will never get back to the Win2K
    gig interface. To fix that set up a static route on the AIX box.

    route add -host 170.29.112.254 -netmask 255.255.255.255 170.29.112.30

    Bruce Zimmer
    Central Data Systems
    (248) 615-4644 (direct)
    (248) 320-1175 (cell)
    bzimmer@centraldata.com

    -----Original Message-----
    From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU] On Behalf Of
    Willeat, Todd
    Sent: Friday, April 25, 2003 12:56 AM
    To: aix-l@Princeton.EDU
    Subject: Re: Strange networking issue

    There shouldn't be any routing since both boxes are on the same
    subnet...
    But, if it helps, here's the traceroute info:

    trying to get source for mhp-aquaman
    source should be 170.29.112.254
    traceroute to mhp-aquaman (170.29.112.30) from 170.29.112.254
    (170.29.112.254), 30 hops max
    outgoing MTU = 1500
     1 * * *
     2 * * *
     3 * * *
     4 * * *
     5 * * *
     6 * * *
     7 * * *
     8 * * *
     9 * * *
    10 * * *
    11 * * *
    12 * * *
    13 * * *
    14 * * *
    15 * * *
    16 * * *
    17 * * *
    18 * * *
    19 * * *
    20 * * *
    21 * * *
    22 * * *
    23 * * *
    24 * * *
    25 * * *
    26 * * *
    27 * * *
    28 * * *
    29 * * *
    30 * * *

    -----Original Message-----
    From: John Jolet [mailto:john.jolet@FXFN.COM]
    Sent: Thursday, April 24, 2003 8:16 PM
    To: aix-l@Princeton.EDU
    Subject: Re: Strange networking issue

    really, really sounds like a routing issue...what does a traceroute from
    the
    aix to win2k box look like?

    -----Original Message-----
    From: Willeat, Todd [mailto:TWilleat@MHP.SMHS.COM]
    Sent: Thursday, April 24, 2003 5:22 PM
    To: aix-l@Princeton.EDU
    Subject: Strange networking issue

    Hi all,

    I have a strange network connectivity issue between a p640 box (AIX
    4.3.3
    ML11) and a Win2K box (SP3). Both boxes are on the same subnet (and same
    switch), but are having problems pinging each other. The Win2K box has
    no
    problem connecting to other boxes, and people can telnet to the AIX box.
    The
    Win2K box has 6 GigE connections in an 802.3ad dynamic team. The p640
    box
    has a SysKonnect dual-port GigE card running in failover mode (as well
    as
    the 2 integrated 10/100s - I have one of these on another subnet and the
    Win2K box can ping that just fine...). Looking at the arp tables shows
    that
    each box knows the other's MAC address. My network admin captured some
    packets between the 2 and this is what we saw:

    Pinging from Win2K to AIX - we saw the packet go from Win2K to AIX, but
    AIX
    did not return any packets.
    Pinging from AIX to Win2K - we saw the packet go from AIX to Win2K and
    saw
    the response packet from Win2K to AIX, but AIX did not receive it
    (request
    timed out).

    It appears that AIX is dropping the packets from this Win2K box for some
    reason (I have several other p640s with the same config that work fine).
    However, when I rebooted the Win2K box (after it locked up...), I could
    then
    ping between the boxes. But this only lasted for a day or so, and now it
    is
    not working again.

    Has anyone seen anything like this before?

                   _______________________________________
                  / Todd A. Willeat \
                 | UNIX Administrator |
                 | Mercy Health Plans |
                 | 425 South Woods Mill Road |
               _ | Chesterfield, Missouri 63017-3492 | _
              / )| (314) 214-2329 / (314) 214-8202 (fax) |( \
             / / | twilleat@mhp.mercy.net | \ \
           _( (_ | http://www.mercyhealthplans.com/ | _) )_
          (((\ \> \/->___________<-=-=-=-=->___________<-\/ </ /)))
          (\\\\ \_/ / \ \_/ ////)
           \ / \ /
            \ _/ \_ /
            / / \ \
           /___/ \___\

    This e-mail contains information which (a) may be PROPRIETARY IN NATURE
    OR
    OTHERWISE PROTECTED BY LAW FROM DISCLOSURE, and (b) is intended only for
    the
    use of the addressee (s) named above. If you are not the addressee, or
    the
    person responsible for delivering this to the addressee (s), you are
    hereby
    notified that reading, copying or distributing this e-mail is
    prohibited. If
    you have received this e-mail in error, please contact the sender
    immediately.


  • Next message: Klaus Oberle: "Re: hdisks missing"

    Relevant Pages

    • Re: Strange networking issue
      ... Subject: Strange networking issue ... Yes same subnet, but what subnet is the 10/100 interface on. ... on the same physical network the packet will never get back to the Win2K ... To fix that set up a static route on the AIX box. ...
      (AIX-L)
    • Re: dhcp diagram
      ... router will add its IP address to that packet. ... every computer receives IP address according to its subnet. ... that's why we need to set up dhcp relay agent on other subnet to ... >> address to the client based on subnet of its dhcp relay agent. ...
      (microsoft.public.win2000.networking)
    • Re: Two ethernet interfaces puzzle
      ... steve wrote: ... > possibly 192.168.2.x) then route the packets from there. ... the ability to "route add" the subnet as a unique route. ... packet after packet, frame by frame. ...
      (alt.os.linux.suse)
    • Re: NTP over redundant peer links, undetected loops
      ... at home is also in the ever-popular 192.168.1.x subnet. ... stack tries to deliver it to some other hotel NAT cesspool customer, ... and the packet never makes it to the VPN. ...
      (comp.protocols.time.ntp)
    • Re: Subnetmask Issues
      ... > another machine which is in the same IP subnet, the router is not involved ... The packet is delivered directly (ie "on the wire" using the ... >> How would the router identify the specific network in which the client ...
      (microsoft.public.windows.server.networking)