Re: [SLE] SAMBA Problem

From: Stephen P. Molnar, Ph.D. (smolnar_at_jadeinc.com)
Date: 07/29/03

  • Next message: Bruce Marshall: "Re: [SLE] Upgrade to 7.2 to SuSE 8.2" 
    To: suse-linux-e@suse.com
Date: Tue, 29 Jul 2003 08:28:35 -0400

    I trust that this is a friendly exchange. I would hate to think that my LAN
    problem is causing dissention.

    A few more observations about the problem.

    Someone suggested that the problem was due to the hub or the cables. I don't
    think that this is the case as my linux machine will access my laptop running
    Win2000pro via the LAN without any problems. However, LinNeighborhood tells
    me that:

    GetSMBShare:smbclient -L INGA -I 10.0.0.3 -U% -W FFC -d3
    Initialising global parameters
    params.c:pm_process() - Processing configuration file "//etc/samba/smb.conf"
    Processing section "[global]"
    added interface ip=10.0.0.2 bcast=10.0.0.255 nmask=255.255.255.0
    Client started (version 2.2.5).
    Connecting to 10.0.0.3 at port 139
    Domain=[FFC] OS=[Windows 5.0] Server=[Windows 2000 LAN Manager]

            Sharename Type Comment
            --------- ---- -------
    Error returning browse list: NT_STATUS_ACCESS_DENIED

            Server Comment
            --------- -------
            INGA

            Workgroup Master
            --------- -------
            FFC INGA

    However, I can mount INGA via SAMBA by executing smbmount in a console.

    Having answered the hardware question. I find that I can ping both ways
    between the linux platform and the laptop.

    If, however I ping the W2000p box (not the laptop) I get:

    abnormal:~ # ping 10.0.0.1
    PING 10.0.0.1 (10.0.0.1) from 10.0.0.2 : 56(84) bytes of data.
    From 10.0.0.2: icmp_seq=1 Destination Host Unreachable
    From 10.0.0.2 icmp_seq=1 Destination Host Unreachable
    From 10.0.0.2 icmp_seq=2 Destination Host Unreachable
    From 10.0.0.2 icmp_seq=3 Destination Host Unreachable

    --- 10.0.0.1 ping statistics ---
    6 packets transmitted, 0 received, +4 errors, 100% loss, time 5028ms
    , pipe 3

    Now if I switch to the W2000p platform pinging either other machine doesn't
    work.

    I am forced to conclude that the problem wies with W2000p on that machine.
    Therefore, I am really at sea.
    On Monday 28 July 2003 06:57 pm, Jim Cunning wrote:
    > Monday July 28 at 4:03pm, Thomas Jones wrote:
    > > On Monday 28 July 2003 12:03, Jim Cunning wrote:
    > > <snip>
    > >
    > > > WRONG!! The netmask shown is surely not the reason why he cannot talk
    > > > to the LAN. I don't know what the solution to Dr. Molnar's problem is,
    > > > but changing to a class A netmask is barking up the wrong tree.
    > > >
    > > > There is absolutely no requirement that a 10.x.x.x network use a class
    > > > A netmask--we use Class C netmasks on such a network with no problems
    > > > whatsoever. Ever since the adoption of CIDR (Classless Inter-Domain
    > > > Routing - see RFC 1519), the concept of Class A, B and C netmasks has
    > > > been obsolete.
    > > >
    > > > Jim
    > >
    > > Jim,
    > >
    > > I understand CIDR....i have a certificate in TCP/IP administration. And
    > > in fact, i did calculate his broadcast address by supernetting his subnet
    > > and using ANDing.
    >
    > If you did, then you must have concluded that the broadcast address he
    > had for the network address and mask was correct.
    > Address: 10.0.0.2, broadcast: 10.0.0.255, and netmask 255.255.255.0 is
    > just fine.
    >
    > > However, what you fail to realize is that for the CIDR implementation to
    > > work there have to be various things to possibly work:
    > >
    > > 1. No routing device between the hosts in question.
    >
    > Dr. Molnar did say it was a small LAN, so that is a safe assumption. He
    > was having trouble with 10.0.0.2 not reaching 10.0.0.1. With ANY netmask
    > we've discussed (class A or C), those two hosts have to be on the same
    > subnet, so routers are irrelevant.
    >
    > > 2. If a routing device does exist in the network, it must be CIDR
    > > compatible.
    >
    > OK, but that isn't an issue because of #1.
    >
    > > 3. If that device is compatible it must be configured to correctly
    > > process the proper routing protocol ---- as in BGP or RIPv2.
    >
    > This doesn't follow at all. There is no need for any routing protocols
    > here for a CIDR compatible router to work. We have a lab with hundreds of
    > IP hosts and many different sorts of routers. None of them has BGP or RIP
    > configured, yet we have many networks: 10.x.x.x/12 and /24, 172.16.x.x/20,
    > 192.168.x.x/24, /28, and /30. All work just fine without routing
    > protocols, just a lot of static routes.
    >
    > > So as you can see, your explanation that CIDR WILL work is incorrectly
    > > stated. Specifically for private/residential networks, the use of BGP
    > > is next to nil.
    >
    > No I don't see, though from my response above I agree that BGP use is low
    > in those networks.
    >
    > > Nevertheless, this is fairly easy to determine.
    > >
    > > All he has to do is ping the IP addy of the other node. Then attempt to
    > > ping an outside entity ----maybe www.google.com - 216.239.37.99 .
    > >
    > > Do so from both nodes.
    > >
    > > If he cannot ping the local node, and the node is in fact accessing the
    > > WAN; than it is the routing at fault. That's all there is to it.
    >
    > I wouldn't be so categorical. First, Dr. Molnar never said he was
    > accessing the WAN from either of the hosts he identified. Second, routing
    > is far from the only possible reason for failure to reach a local host.
    >
    > Earlier in this thread, Dr. Molnar said that ifconfig only showed two
    > interfaces: eth0 and lo. If you look at his output from eth0
    >
    > > eth0 Link encap:Ethernet HWaddr 00:A0:CC:78:C9:1F
    > > inet addr:10.0.0.2 Bcast:10.0.0.255 Mask:255.255.255.0
    > > inet6 addr: fe80::2a0:ccff:fe78:c91f/10 Scope:Link
    > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
    > > RX packets:0 errors:0 dropped:0 overruns:0 frame:0
    > > TX packets:133 errors:0 dropped:0 overruns:0 carrier:0
    > > collisions:0 txqueuelen:100
    > > RX bytes:0 (0.0 b) TX bytes:6790 (6.6 Kb)
    > > Interrupt:11 Base address:0x7000
    >
    > you'll see that 133 packets have been transmitted, but NONE has been
    > received. He also said that tcpdump reported no traffic. From this I
    > would suspect: ethernet cable, NIC (in either host on the LAN), or
    > ethernet hub/switch.
    >
    > BTW, when mount fails to reach a host (using smbfs, at least), perhaps for
    > one of the faults I listed above, it reports
    >
    > test:~ # mount -t smbfs //testhost12/N /mnt
    > Error connecting to 10.1.1.12 (No route to host)
    > 24534: Connection to testhost failed
    > SMB connection failed
    >
    > The "No route to host" is a bit of a red herring, because there was no
    > routing problem in my test, just a non-responsive host (because it wasn't
    > there).
    >
    > Jim 

    -- 
Stephen P. Molnar, Ph.D.			Life is a fuzzy set
Foundation for Chemistry			Stochastic and multivariant
http://web.jadeinc.com/FoundationChem
-- 
Check the headers for your unsubscription address
For additional commands send e-mail to suse-linux-e-help@suse.com
Also check the archives at http://lists.suse.com
Please read the FAQs: suse-linux-e-faq@suse.com
  • Next message: Bruce Marshall: "Re: [SLE] Upgrade to 7.2 to SuSE 8.2"

    Relevant Pages

    • Re: OpenVPN configuration problem
      ... > - when i ping from the client a host on server's LAN, ...
      (comp.os.linux.networking)
    • Re: ICS questions and confusion
      ... >>> has to be another subnet altogether. ... WHY does the LAN connection that connects my router to the ... >>host have to be on a different subnet than the 192.168.0.x one? ... >>where is that documented in all the how-to's on ICS, ...
      (microsoft.public.windowsxp.network_web)
    • Re: Setting routes w/ set next hop verify-availability in IOS 12.2
      ... route for a particular host pointing to 10.100.20.5. ... particular host if there is a good ping to that specific outbound ... Will the host then just use whatever route is in the MSFC2's routing ... connections and 7200 routers running OSPF without getting into BGP. ...
      (comp.dcom.sys.cisco)
    • Re: router causing ssh etc. slowdown?
      ... >> port on their public IP from a host on their own LAN. ... Let's say host A starts a TCP connection from its ... The router sends the SYN packet to ...
      (Fedora)
    • Re: how to "join" LAN with plip link?
      ... PLIP howto) about a real case on how I have made a plip connected host ... looking like on the LAN. ... Then I test the proxy-arp by ... LAN (or client), or you need to configure the server to do 'proxy-ARP'. ...
      (comp.os.linux.networking)