[Beowulf] noob Root-NFS question

Michael Di Domenico mdidomenico4 at gmail.com
Sat May 9 18:51:25 PDT 2009


Just a shot in the dark, but try exporting /diskless along with
/diskless/192.168.2.11

If i recall correctly NFS-Root in the kernel will try to mount
/diskless and then look for a directory with the corresponding IP
address

(i could be wrong though, i haven't done it in awhile)

On Sun, May 3, 2009 at 6:25 PM, Cocoy Dayao <cocoy.dayao at gmail.com> wrote:
> Dear list,
> I'm not sure if this is the right forum for this. But anyway here goes. i
> hope you could help me out. So I wanted to play around with building my own
> Beowulf Cluster. Yes, I know there are easier ways to do this. Automated
> tools like Caos Linux, etc. etc. Anyway, I wanted to use gentoo and learn
> from the ground up. I wanted a diskless client. So I booted a box via pxe,
> tftp and dhcp. it boots--- but can't seem to find NFS.
> And yes, i've googled. none have worked. I don't know what I'm missing.
> I get this:
> rpcbind: server 192.168.2.1 not responding, timed out
> Root-NFS: Unable to get nfsd port number from server, using default
> Looking up port of RPC 100005/1 on 192.168.2.1
> rpcbind: server 192.168.2.1 not responding, timed out
> Root-NFS: Unable to get mountd port number from server, using default
> Root-NFS: Server returned error -5 while mounting /diskless/192.168.2.11
> VFS: Unable to mount root fs via NFS, tryi9ng floppy.
> VFS: Cannot open root device "nfs" or unknown-block(2,0)
> Kernel Panic - not syncing: VFS unable to mount root fs on
> unknown-block(2,0)
> Both client and server have root nfs turned on it their respective kernel.
> i have turned off the firewall on the server and still get the same error.
> pxelinux.cfg is this:
> DEFAULT /kernel8
> APPEND root=/dev/nfs rw
> nfsroot=192.168.2.1:/diskless/192.168.2.11 init=sbin/init
> this is rpcinfo:
> talon dhcp # rpcinfo -p 192.168.2.1
>    program vers proto   port
>     100000    4   tcp    111  portmapper
>     100000    3   tcp    111  portmapper
>     100000    2   tcp    111  portmapper
>     100000    4   udp    111  portmapper
>     100000    3   udp    111  portmapper
>     100000    2   udp    111  portmapper
>     100024    1   udp  45975  status
>     100024    1   tcp  57882  status
>     100005    1   udp  57290  mountd
>     100005    1   tcp  50765  mountd
>     100005    2   udp  57290  mountd
>     100005    2   tcp  50765  mountd
>     100005    3   udp  57290  mountd
>     100005    3   tcp  50765  mountd
>     100003    2   udp   2049  nfs
>     100003    3   udp   2049  nfs
>     100021    1   udp  57739  nlockmgr
>     100021    3   udp  57739  nlockmgr
>     100021    4   udp  57739  nlockmgr
>     100021    1   tcp  45392  nlockmgr
>     100021    3   tcp  45392  nlockmgr
>     100021    4   tcp  45392  nlockmgr
>     100003    2   tcp   2049  nfs
>     100003    3   tcp   2049  nfs
> ps -aef | grep rpc is this:
> alon conf.d # ps -aef | grep rpc
> root      1101     2  0 18:14 ?        00:00:00 [rpciod/0]
> root      1102     2  0 18:14 ?        00:00:00 [rpciod/1]
> root      8332     1  0 18:15 ?        00:00:00 /sbin/rpcbind
> nobody    8356     1  0 18:15 ?        00:00:00 /sbin/rpc.statd --no-notify
> root      8379     1  0 18:15 ?        00:00:00 /usr/sbin/rpc.mountd
> root      8587  8560  0 18:22 pts/0    00:00:00 grep --colour=auto rpc
> tcpdump:
> 17), length 57) master.talon.11978 > node01.talon.57100: UDP, length 29
> 17:39:47.683582 IP (tos 0x0, ttl 64, id 53556, offset 0, flags [DF], proto
> UDP (17), length 52) master.talon.11974 > node01.talon.57099: UDP, length 24
> 17:39:48.451700 IP (tos 0x0, ttl 64, id 54326, offset 0, flags [DF], proto
> UDP (17), length 57) master.talon.11976 > node01.talon.57100: UDP, length 29
> 17:39:49.665576 IP (tos 0x0, ttl 64, id 63547, offset 0, flags [DF], proto
> UDP (17), length 57) master.talon.11978 > node01.talon.57100: UDP, length 29
> 17:39:49.762700 IP (tos 0x0, ttl 64, id 55637, offset 0, flags [DF], proto
> UDP (17), length 57) master.talon.11975 > node01.talon.57100: UDP, length 29
> 17:39:50.661534 arp who-has node01.talon tell master.talon
> 17:39:51.662530 arp who-has node01.talon tell master.talon
> 17:39:52.401575 IP (tos 0x0, ttl 64, id 58276, offset 0, flags [DF], proto
> UDP (17), length 57) master.talon.11977 > node01.talon.57100: UDP, length 29
> 17:39:52.662526 arp who-has node01.talon tell master.talon
> 17:39:54.471660 arp who-has node01.talon tell master.talon
> arp--- that's the point where kernel panic occurs.
> this is my /etc/exports file:
> #/etc/exports: NFS file systems being exported.  See exports(5).
> /diskless/192.168.2.11  *(rw,no_root_squash,no_all_squash,no_subtree_check)
> /opt    192.168.2.0/24(ro,no_root_squash,no_all_squash,no_subtree_check)
> /usr    192.168.2.0/24(ro,no_root_squash,no_all_squash,no_subtree_check)
> /home   192.168.2.0/24(rw,no_root_squash,no_all_squash,no_subtree_check)
> /var/log
>  192.168.2.11(rw,no_root_squash,no_all_squash,no_subtree_check)
> my dhcp configuration is this:
> # my dhcpd.conf for diskless clients
> allow booting;
> #allow bootp;
> #tftp
> next-server 192.168.2.1;
> #option root-path "/diskless/192.168.2.11";
> option space PXE;
> option PXE.mtftp-ip               code 1 = ip-address;
> option PXE.mtftp-cport            code 2 = unsigned integer 16;
> option PXE.mtftp-sport            code 3 = unsigned integer 16;
> option PXE.mtftp-tmout            code 4 = unsigned integer 8;
> option PXE.mtftp-delay            code 5 = unsigned integer 8;
> option PXE.discovery-control      code 6 = unsigned integer 8;
> option PXE.discovery-mcast-addr   code 7 = ip-address;
> subnet 192.168.2.0 netmask 255.255.255.128 {
>         range 192.168.2.11 192.168.2.20;
>         option domain-name-servers 192.168.2.1;
>         option domain-name "talon";
>         option routers 192.168.2.1;
>         option broadcast-address 192.168.2.195;
>         option root-path "192.168.2.1:/diskless/192.168.2.11";
>         default-lease-time 600;
>         max-lease-time 7200;
>         next-server 192.168.2.1;
>         class "pxeclient" {
>                 match if substring (option vendor-class-identifier, 0, 9) =
> "PXEClient";
>                 vendor-option-space PXE;
>                 option PXE.mtftp-ip 0.0.0.0;
>                 #option PXE.mtftp-ip 192.168.2.1;
>                 filename "pxelinux.0";
>         }
>       # host decleration for diskless node
>
>        host node01.talon {
>                                 hardware ethernet 00:1c:c0:4f:bd:e1;
>                                 fixed-address 192.168.2.11;
>         }
> }
> I also tried this: i used the same cable and attached it to my mac and
> mounted /diskless/192.168.2.11 nfs share which points to the diskless'
> client's root. i was able to mount it. So i know nfs works. And yes i made
> sure to turn off the mac's wireless... so only ethernet was plugged in.
> What did I miss?
> Appreciate, your advice
> Cocoy
> www.twitter.com/cocoy
> "People who are really serious about software should make their own
> hardware" -- Alan Kay
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing
> To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf
>
>




More information about the Beowulf mailing list