NFS sharing not working 2008 - Mandriva

This is a discussion on NFS sharing not working 2008 - Mandriva ; There is an item in errata for 2008 that say the share tool in MCC wont see shares unless: 1) you have fully named your computers (ie. office.home.network guest.home.network -etc.) and then identify the possible share in hosts (ie 192.168.2.121 ...

+ Reply to Thread
Results 1 to 19 of 19

Thread: NFS sharing not working 2008

  1. NFS sharing not working 2008

    There is an item in errata for 2008 that say the share tool in MCC wont
    see shares unless: 1) you have fully named your computers (ie.
    office.home.network guest.home.network -etc.) and then identify the
    possible share in hosts (ie 192.168.2.121 guest.home.network guest )
    however this has not helped me at all.

    I can see the downstairs computer running 2007.1 but the two other
    computers do no see each others shares.

    I would be grateful for any advise, although I think that not
    networking is well appalling for mandy and the MCC app should work, as
    in omg. As in ---???

  2. Re: NFS sharing not working 2008

    On Sun, 11 Nov 2007 12:28:48 -0400, Marlene wrote:
    >
    > I can see the downstairs computer running 2007.1 but the two other
    > computers do no see each others shares.


    Every computer needs to know the FQDN names and ips of eveyone they
    want to communicate. This is done in the /etc/hosts files.
    I have the same /etc/hosts file on all systems. Snippet follows:

    $ head /etc/hosts
    127.0.0.1 localhost
    192.168.1.31 wb7_0.home.invalid wb7_0
    192.168.1.11 fw.home.invalid fw
    192.168.1.12 wb1.home.invalid wb1
    192.168.1.1 cable_modem cm
    192.168.1.130 wb.home.invalid wb
    192.168.1.131 beta.home.invalid beta
    192.168.1.132 kubu7.home.invalid kubu7
    192.168.1.133 fc7.home.invalid fc7
    192.168.1.134 fc8.home.invalid fc8

    Sorry I can not be of anymore help on shares.

  3. Re: NFS sharing not working 2008

    Bit Twister wrote:
    > On Sun, 11 Nov 2007 12:28:48 -0400, Marlene wrote:
    >> I can see the downstairs computer running 2007.1 but the two other
    >> computers do no see each others shares.

    >
    > Every computer needs to know the FQDN names and ips of eveyone they
    > want to communicate. This is done in the /etc/hosts files.
    > I have the same /etc/hosts file on all systems. Snippet follows:
    >
    > $ head /etc/hosts
    > 127.0.0.1 localhost
    > 192.168.1.31 wb7_0.home.invalid wb7_0
    > 192.168.1.11 fw.home.invalid fw
    > 192.168.1.12 wb1.home.invalid wb1
    > 192.168.1.1 cable_modem cm
    > 192.168.1.130 wb.home.invalid wb
    > 192.168.1.131 beta.home.invalid beta
    > 192.168.1.132 kubu7.home.invalid kubu7
    > 192.168.1.133 fc7.home.invalid fc7
    > 192.168.1.134 fc8.home.invalid fc8
    >
    > Sorry I can not be of anymore help on shares.



    The one "downstairs" (which is a 2007.1) does not have a name does not
    have anything it its hosts file, except 127.0.0.1 - it can be seen and
    its shares mounted. I named (the 2008 boxen) only in an attempt (I
    have never given fqdn, other names other than default localhost) to
    allow their shares to be seen. Until 2008, all boxen, shared alike and
    mounted alike.

    At this moment only the 2007.1 box has shares that can be seen or mounted.

  4. Re: NFS sharing not working 2008

    Bit Twister wrote:
    > On Sun, 11 Nov 2007 12:28:48 -0400, Marlene wrote:
    >> I can see the downstairs computer running 2007.1 but the two other
    >> computers do no see each others shares.

    >
    > Every computer needs to know the FQDN names and ips of eveyone they
    > want to communicate. This is done in the /etc/hosts files.
    > I have the same /etc/hosts file on all systems. Snippet follows:
    >
    > $ head /etc/hosts
    > 127.0.0.1 localhost
    > 192.168.1.31 wb7_0.home.invalid wb7_0
    > 192.168.1.11 fw.home.invalid fw
    > 192.168.1.12 wb1.home.invalid wb1
    > 192.168.1.1 cable_modem cm
    > 192.168.1.130 wb.home.invalid wb
    > 192.168.1.131 beta.home.invalid beta
    > 192.168.1.132 kubu7.home.invalid kubu7
    > 192.168.1.133 fc7.home.invalid fc7
    > 192.168.1.134 fc8.home.invalid fc8
    >
    > Sorry I can not be of anymore help on shares.



    Heres another thing, I just plugged my camera in and digiKam refuses to
    load the camera. Using system:/media/camera/ I get "cannot lock" errors.

    I feel like this is related but I don't know.

  5. Re: NFS sharing not working 2008

    On Sun, 11 Nov 2007 14:19:24 -0400, Marlene wrote:
    >
    > Heres another thing, I just plugged my camera in and digiKam refuses to
    > load the camera. Using system:/media/camera/ I get "cannot lock" errors.
    >
    > I feel like this is related but I don't know.


    I do not think so either

    Now, you posted some doc info indicating FQDN was required.
    Then that is what you set first. I know postfix and cups complain if
    no FQDN for node.

    Just because it worked on the old release, does not mean system will continue
    to be tollerent of old settings. Upgrades instead of clean installs
    can have problem with old config file settings.

    Newer releases of software may do more checking/adhearing to RFCs.

    Read the last four lines of this reply, I'll wait. . . . . . . . . . . . .



    Based on all the info you have provided so far, my recommendation,
    pick some domain name for your LAN.

    Put ".invalid" on the end so if anything escapes your LAN, possible
    upline victims will not be bothered or their filters will already be
    trashing *.invalid messages.

    You pick names for every node,
    put everyone's "ip FQDN alias" in a hosts file,

    For each node, set node's FQDN in /etc/sysconfig/network,
    copy master /etc/hosts node onto the node
    do a
    chkconfig --del avahi-daemon
    shutdown -h

    Once all nodes are down, power down router, wait 30 seconds by watch/clock
    Bring node connected to internet device up, then each node.

    Assuming no typing errors, everything should come up.
    You then
    ping -c1 each_node
    to check each node/ip to verify it is who you think it is.

    /etc/sysconfig/network contains linux node name used during boot up.
    I suggest your's looks like mine with whatever you pick for node and
    domain name.

    $ cat /etc/sysconfig/network
    NETWORKING_IPV6=no # speedup DNS lookups since ipv6 not used everywhere
    NOZEROCONF=yes # doze ip lookups not needed if router is dead.
    NEEDHOSTNAME=no # I'll use my own hostname
    NETWORKING=yes
    HOSTNAME=wb.home.invalid

    Note/Caution:
    When you change node name/ip address your gui desktop
    manager may freeze. Other deamon/services may need those so a reboot
    takes care of that. The exception is cups, it needs the new node name
    fix in it's config file and a service cups restart.

    You indicated one ip addy of 127.0.0.1. Guessing it is using dhcp
    to get an address or dialup. If using a fancy adsl/cable/fios router, you
    may be able to set the LAN's domain name.

    The dhcp server/router is required to resolve your ip addy for any node
    using dhcp so that node.domain would not be in everyone elses /etc/hosts.

    Repeat host file snippet to be used as a guide. Windoze also has a
    etc\hosts file somewhere

    $ head /etc/hosts
    127.0.0.1 localhost
    192.168.1.31 wb7_0.home.invalid wb7_0
    192.168.1.11 fw.home.invalid fw
    192.168.1.12 wb1.home.invalid wb1
    192.168.1.1 cable_modem cm
    192.168.1.130 wb.home.invalid wb
    192.168.1.131 beta.home.invalid beta
    192.168.1.132 kubu7.home.invalid kubu7
    192.168.1.133 fc7.home.invalid fc7
    192.168.1.134 fc8.home.invalid fc8

    --
    The warranty and liability expired as you read this message.
    If the above breaks your system, it's yours and you keep both pieces.
    Practice safe computing. Backup the file before you change it.
    Do a, man command_here or cat command_here, before using it.

  6. Re: NFS sharing not working 2008

    On Sun, 11 Nov 2007 14:11:13 -0400, Marlene wrote:
    >
    > The one "downstairs" (which is a 2007.1) does not have a name does not
    > have anything it its hosts file, except 127.0.0.1 - it can be seen and
    > its shares mounted. I named (the 2008 boxen) only in an attempt (I
    > have never given fqdn, other names other than default localhost) to
    > allow their shares to be seen. Until 2008, all boxen, shared alike and
    > mounted alike.
    >
    > At this moment only the 2007.1 box has shares that can be seen or mounted.


    Sorry, when you mentioned shares in the body of the text, I assumed
    windoze shares. Just saw you subject line and can help a little more.

    Possible problem areas for NFS are:
    serving share's /etc/exports does not have entry of clients.
    Clients do not have share server line in /etc/hosts for mounting
    /etc/hosts.allow and /etc/hosts.deny, firewalls helping you into the ditch

    file locking/nfs deamon/services not running.

    and if all else fails, click up a terminal
    su - root
    tail -f /var/log/messages

    and try your mount from another terminal or whatever you do to see the share.

    tools of interest are (depending if installed or still used/available)
    exportfs -v
    rpcinfo -p
    showmount
    nfsddebug, nfsstat, nfsdebug

  7. Re: NFS sharing not working 2008

    on Monday 12 November 2007 03:40
    in the Usenet newsgroup alt.os.linux.mandriva
    Bit Twister wrote:

    > On Sun, 11 Nov 2007 12:28:48 -0400, Marlene wrote:
    >>
    >> I can see the downstairs computer running 2007.1 but the two other
    >> computers do no see each others shares.


    [snip]

    Sorry to jump in mid thread, but look at this.

    http://wiki.mandriva.com/en/Releases...shares.22_tool

    --
    sig goes here...
    Peter D.

  8. Re: NFS sharing not working 2008

    On Mon, 12 Nov 2007 09:25:44 +1100, Peter D. wrote:

    > Sorry to jump in mid thread, but look at this.


    > http://wiki.mandriva.com/en/Releases...shares.22_tool
    >


    But notice that the workaround stills want FQDNs

    But you would think the
    $ ls -al /usr/lib/libDrakX/fs/mount.pm
    -rw-r--r-- 1 root root 7186 2007-11-02 15:33 /usr/lib/libDrakX/fs/mount.pm
    copy would have the fix.

    I always assume user has all updates installed.


  9. Re: NFS sharing not working 2008

    Peter D. wrote:
    > on Monday 12 November 2007 03:40
    > in the Usenet newsgroup alt.os.linux.mandriva
    > Bit Twister wrote:
    >
    >> On Sun, 11 Nov 2007 12:28:48 -0400, Marlene wrote:
    >>> I can see the downstairs computer running 2007.1 but the two other
    >>> computers do no see each others shares.

    >
    > [snip]
    >
    > Sorry to jump in mid thread, but look at this.
    >
    > http://wiki.mandriva.com/en/Releases...shares.22_tool
    >

    thanks peter i reference that in my first post

  10. Re: NFS sharing not working 2008

    Bit Twister wrote:
    > On Sun, 11 Nov 2007 14:19:24 -0400, Marlene wrote:
    >> Heres another thing, I just plugged my camera in and digiKam refuses to
    >> load the camera. Using system:/media/camera/ I get "cannot lock" errors.
    >>
    >> I feel like this is related but I don't know.

    >
    > I do not think so either
    >
    > Now, you posted some doc info indicating FQDN was required.


    No the errata notice -I referenced-and peter refers to later in this
    thread, says the MCC tool does not mount nfs shares, and that a
    "workaround" is to name (fqdn) the host computer and to place that fqdn
    in /etc/hosts on the box that wants to use its share









    > Then that is what you set first. I know postfix and cups complain if
    > no FQDN for node.
    >
    > Just because it worked on the old release, does not mean system will continue
    > to be tollerent of old settings. Upgrades instead of clean installs
    > can have problem with old config file settings.
    >
    > Newer releases of software may do more checking/adhearing to RFCs.
    >
    > Read the last four lines of this reply, I'll wait. . . . . . . . . . . . .
    >


    Im sorry if im just being sensitive but this implies that I refused to
    name my boxen and refused to place the required lines in /etc/hosts,
    in actual fact I reference doing as asked by the errata note, and added
    in further explanation or experts delectation, that although not
    named(fqdn), and with no reference in /etc/hosts the 2008 box happily
    detected and mounted my as yet [!! :-) ] un-up-graded 2007.1 box's shares.

    My though, in making this info available, was that the problem is not
    how the receiving computer see's offered shares but how the share
    offering (2008) boxen make their offer.






    >
    >
    > Based on all the info you have provided so far, my recommendation,
    > pick some domain name for your LAN.


    as i said in the first post that is what i did when i read the errata.
    >
    > Put ".invalid" on the end so if anything escapes your LAN, possible
    > upline victims will not be bothered or their filters will already be
    > trashing *.invalid messages.
    >


    having learned much from you in the past, I had already used .invalid
    even tho you had suggested the reason for doing so was unecessarily
    complex or obscure

    > You pick names for every node,
    > put everyone's "ip FQDN alias" in a hosts file,
    >
    > For each node, set node's FQDN in /etc/sysconfig/network,
    > copy master /etc/hosts node onto the node
    > do a
    > chkconfig --del avahi-daemon
    > shutdown -h
    >
    > Once all nodes are down, power down router, wait 30 seconds by watch/clock
    > Bring node connected to internet device up, then each node.
    >
    > Assuming no typing errors, everything should come up.
    > You then
    > ping -c1 each_node
    > to check each node/ip to verify it is who you think it is.
    >
    > /etc/sysconfig/network contains linux node name used during boot up.
    > I suggest your's looks like mine with whatever you pick for node and
    > domain name.
    >
    > $ cat /etc/sysconfig/network
    > NETWORKING_IPV6=no # speedup DNS lookups since ipv6 not used everywhere
    > NOZEROCONF=yes # doze ip lookups not needed if router is dead.
    > NEEDHOSTNAME=no # I'll use my own hostname
    > NETWORKING=yes
    > HOSTNAME=wb.home.invalid
    >
    > Note/Caution:
    > When you change node name/ip address your gui desktop
    > manager may freeze. Other deamon/services may need those so a reboot
    > takes care of that. The exception is cups, it needs the new node name
    > fix in it's config file and a service cups restart.
    >
    > You indicated one ip addy of 127.0.0.1. Guessing it is using dhcp
    > to get an address or dialup. If using a fancy adsl/cable/fios router, you
    > may be able to set the LAN's domain name.


    nope that was there 127.0.0.1 merely because I did not choose a fqdn
    when i installed _ Ihave doen reboots with 127.0.0.1 commented out and
    left in. It now is correctly set to the name of the box it resides on
    so that it looks like this
    127.0.0.1 thisbox.thisnet.invalid



    >
    > The dhcp server/router is required to resolve your ip addy for any node
    > using dhcp so that node.domain would not be in everyone elses /etc/hosts.
    >
    > Repeat host file snippet to be used as a guide. Windoze also has a
    > etc\hosts file somewhere
    >
    > $ head /etc/hosts
    > 127.0.0.1 localhost
    > 192.168.1.31 wb7_0.home.invalid wb7_0
    > 192.168.1.11 fw.home.invalid fw
    > 192.168.1.12 wb1.home.invalid wb1
    > 192.168.1.1 cable_modem cm
    > 192.168.1.130 wb.home.invalid wb
    > 192.168.1.131 beta.home.invalid beta
    > 192.168.1.132 kubu7.home.invalid kubu7
    > 192.168.1.133 fc7.home.invalid fc7
    > 192.168.1.134 fc8.home.invalid fc8
    >


  11. Re: NFS sharing not working 2008

    Bit Twister wrote:
    > On Mon, 12 Nov 2007 09:25:44 +1100, Peter D. wrote:
    >
    >> Sorry to jump in mid thread, but look at this.

    >
    >> http://wiki.mandriva.com/en/Releases...shares.22_tool
    >>

    >
    > But notice that the workaround stills want FQDNs
    >
    > But you would think the
    > $ ls -al /usr/lib/libDrakX/fs/mount.pm
    > -rw-r--r-- 1 root root 7186 2007-11-02 15:33 /usr/lib/libDrakX/fs/mount.pm
    > copy would have the fix.
    >
    > I always assume user has all updates installed.
    >



    as elsewher expalined I used the fqdn workarouind -with no result
    it was where i started


  12. Re: NFS sharing not working 2008

    On Sun, 11 Nov 2007 19:49:47 -0400, Marlene wrote:
    > Bit Twister wrote:
    >>
    >> Now, you posted some doc info indicating FQDN was required.

    >
    > No the errata notice -I referenced-and peter refers to later in this
    > thread, says the MCC tool does not mount nfs shares, and that a
    > "workaround" is to name (fqdn) the host computer and to place that fqdn
    > in /etc/hosts on the box that wants to use its share


    Yep, FQDN you mentioned.
    >>
    >> Read the last four lines of this reply, I'll wait. . . . . . . . . . . . .
    >>

    >
    > Im sorry if im just being sensitive but this implies that I refused to
    > name my boxen and refused to place the required lines in /etc/hosts,
    > in actual fact I reference doing as asked by the errata note, and added
    > in further explanation or experts delectation, that although not
    > named(fqdn), and with no reference in /etc/hosts the 2008 box happily
    > detected and mounted my as yet [!! :-) ] un-up-graded 2007.1 box's shares.


    Sorry, no need to be sensitive. As I misunderstood your thread so far,
    I was not sure you FQDN every node and set hosts file.

    > My though, in making this info available, was that the problem is not
    > how the receiving computer see's offered shares but how the share
    > offering (2008) boxen make their offer.


    Ok, Sorry, My Fault. Thought you wanted to fix the problem and was
    offering information.

    >> Based on all the info you have provided so far, my recommendation,
    >> pick some domain name for your LAN.

    >
    > as i said in the first post that is what i did when i read the errata.


    Yes, but do bear in mind, I write for the lurkers and for someone
    using a search engine to see what you did to fix the same problem on
    their system.

    >
    > having learned much from you in the past, I had already used .invalid
    > even tho you had suggested the reason for doing so was unecessarily
    > complex or obscure


    What can I say, pointing a newbie to
    http://www.rfc-editor.org/rfc/rfc2606.txt would be a bit cold in my opinion.


    >>
    >> You indicated one ip addy of 127.0.0.1. Guessing it is using dhcp
    >> to get an address or dialup. If using a fancy adsl/cable/fios router, you
    >> may be able to set the LAN's domain name.

    >
    > nope that was there 127.0.0.1 merely because I did not choose a fqdn
    > when i installed _ Ihave doen reboots with 127.0.0.1 commented out and
    > left in. It now is correctly set to the name of the box it resides on
    > so that it looks like this
    > 127.0.0.1 thisbox.thisnet.invalid


    Well, that bites. Assuming not using dhcp for ip addrdess, your
    /etc/hosts should look some what like

    cat /etc/hosts
    127.0.0.1 localhost
    192.168.1.31 thisbox.thisnet.invalid thisbox
    192.168.1.20 thatbox.thisnet.invalid thatbox
    192.168.1.110 anotherbox.thisnet.invalid anotherbox


    Trust/humor me, first line should be
    127.0.0.1 localhost
    or 127.0.0.1 localhost.domain localhost

    As you can see from the following snippet, some local connections within the
    node use localhost

    $ netstat -p

    Proto Recv-Q Send-Q Local Address Foreign Address State
    tcp 0 localhost:6013 localhost:44893 ESTABLISHED
    tcp 0 localhost:37454 localhost:6012 ESTABLISHED
    tcp 0 localhost:6015 localhost:49874 ESTABLISHED
    tcp 0 localhost:6014 localhost:58131 ESTABLISHED
    tcp 0 wb.home.invalid:50936 wb.home.invalid:ssh ESTABLISHED
    tcp 0 localhost:x11-ssh-offset localhost:50019 ESTABLISHED
    tcp 32 wb.home.invalid:50385 news.verizon.net:nntp CLOSE_WAIT

  13. Re: NFS sharing not working 2008

    on Monday 12 November 2007 09:41
    in the Usenet newsgroup alt.os.linux.mandriva
    Bit Twister wrote:

    > On Mon, 12 Nov 2007 09:25:44 +1100, Peter D. wrote:
    >
    >> Sorry to jump in mid thread, but look at this.

    >
    >>

    http://wiki.mandriva.com/en/Releases...shares.22_tool
    >>

    >
    > But notice that the workaround stills want FQDNs


    Yes. There is a danger in jumping into the middle of a thread
    like I did - it is very possible that I am rehashing solved problems.

    > But you would think the
    > $ ls -al /usr/lib/libDrakX/fs/mount.pm
    > -rw-r--r-- 1 root root 7186 2007-11-02 15:33
    > /usr/lib/libDrakX/fs/mount.pm
    > copy would have the fix.


    I hope so.

    for reference;

    rpm -qf /usr/lib/libDrakX/fs/mount.pm
    drakxtools-backend-10.4.239-1mdv2008.0

    > I always assume user has all updates installed.


    Sadly, that is not always a valid assumption.


    --
    sig goes here...
    Peter D.

  14. Re: NFS sharing not working 2008

    Peter D. wrote:
    > on Monday 12 November 2007 09:41
    > in the Usenet newsgroup alt.os.linux.mandriva
    > Bit Twister wrote:
    >
    >> On Mon, 12 Nov 2007 09:25:44 +1100, Peter D. wrote:
    >>
    >>> Sorry to jump in mid thread, but look at this.

    > http://wiki.mandriva.com/en/Releases...shares.22_tool
    >> But notice that the workaround stills want FQDNs

    >
    > Yes. There is a danger in jumping into the middle of a thread
    > like I did - it is very possible that I am rehashing solved problems.
    >
    >> But you would think the
    >> $ ls -al /usr/lib/libDrakX/fs/mount.pm
    >> -rw-r--r-- 1 root root 7186 2007-11-02 15:33
    >> /usr/lib/libDrakX/fs/mount.pm
    >> copy would have the fix.

    >
    > I hope so.
    >
    > for reference;
    >
    > rpm -qf /usr/lib/libDrakX/fs/mount.pm
    > drakxtools-backend-10.4.239-1mdv2008.0
    >
    >> I always assume user has all updates installed.

    >
    > Sadly, that is not always a valid assumption.
    >
    >

    Peter thank you again, I do keep my system as up to date as possible, I
    have done several updates , including yesterday.

    rpm -qf /usr/lib/libDrakX/fs/mount.pm

    also produces the following on my machine

    drakxtools-backend-10.4.239-1mdv2008.0

    but the problem is not solved, the MCC tool for nfs shares does not
    detect, and does not of course mount available shares from either 2008
    (fully updated) installed box to the other 2008 installed box. Both
    have fqdm, both have /etc/hosts files correctly spelled lines,
    /etc/exports on both as written by MCC tools, both have their fqdn
    names set correctly at boot. This is as as much as all documentation
    about the MCC tool problem, and all suggestions from here, applied.

    I will add again that the lone box in the network running 2007.1 (all
    updates) does offer shares (nfs) and can be mounted on both 2008 boxes,
    but the 2007.1 machine does not see the shares offered by either 2008 box.

  15. Re: NFS sharing not working 2008

    On Mon, 12 Nov 2007 09:18:04 -0400, Marlene wrote:
    >
    > I will add again that the lone box in the network running 2007.1 (all
    > updates) does offer nfs shares and can be mounted on both 2008 boxes,
    > but the 2007.1 machine does not see the shares offered by either 2008 box.


    If you were to provide actual config file contents someone might see
    what needs correcting.

    From one 2008 box show us results from these comands
    hostname
    exportfs -v
    cat /etc/exports
    cat /etc/hosts
    cat /etc/hosts.allow
    cat /etc/hosts.deny

    From the 2007.1 box:
    hostname
    exportfs -v
    cat /etc/exports
    cat /etc/fstab
    cat /etc/hosts
    cat /etc/hosts.allow
    cat /etc/hosts.deny

  16. Re: NFS sharing not working 2008

    Marlene wrote:
    > There is an item in errata for 2008 that say the share tool in MCC wont
    > see shares unless: 1) you have fully named your computers (ie.
    > office.home.network guest.home.network -etc.) and then identify the
    > possible share in hosts (ie 192.168.2.121 guest.home.network guest )
    > however this has not helped me at all.
    >
    > I can see the downstairs computer running 2007.1 but the two other
    > computers do no see each others shares.
    >
    > I would be grateful for any advise, although I think that not
    > networking is well appalling for mandy and the MCC app should work, as
    > in omg. As in ---???


    Have you checked /etc/exports, /etc/hosts.deny, and /etc/hosts.allow?
    Could it be a firewall issue? Maybe check the firewall log files?

    G'luck....

    --
    PLEASE post a SUMMARY of the answer(s) to your question(s)!
    Show Windows & Gates to the exit door.
    Unless otherwise noted, the statements herein reflect my personal
    opinions and not those of any organization with which I may be affiliated.

  17. Re: NFS sharing not working 2008

    Marlene wrote:

    > drakxtools-backend-10.4.239-1mdv2008.0
    >
    > but the problem is not solved, the MCC tool for nfs shares does not
    > detect, and does not of course mount available shares from either 2008
    > (fully updated) installed box to the other 2008 installed box. Both
    > have fqdm, both have /etc/hosts files correctly spelled lines,
    > /etc/exports on both as written by MCC tools, both have their fqdn
    > names set correctly at boot. This is as as much as all documentation
    > about the MCC tool problem, and all suggestions from here, applied.
    >
    > I will add again that the lone box in the network running 2007.1 (all
    > updates) does offer shares (nfs) and can be mounted on both 2008 boxes,
    > but the 2007.1 machine does not see the shares offered by either 2008 box.


    I know it doesn't help directly, but nfs works fine on my 2008 machines,
    whats more I don't have fqd names for any of the machines on the network,
    which is mixed 2008 and 2007.1, or a list of all hosts in /etc/hosts. I use
    a linux box as a router/gateway/dns server and all boxes on the network use
    this box for their host info. All boxes do have fixed ip addresses and
    single word host names. I have no firewalls running on the internal network
    and there are no entrys in any /etc/hosts.allow or /etc/hosts.deny.

    If you do as BT suggested and post the contents of the relevant config files
    I will compare to mine.

    --
    sid
    RLU 300284
    Mandriva 2008 X86_64
    2.6.23

  18. SOLVED!! NFS sharing not working 2008



    Marlene wrote:
    > There is an item in errata for 2008 that say the share tool in MCC wont
    > see shares unless: 1) you have fully named your computers (ie.
    > office.home.network guest.home.network -etc.) and then identify the
    > possible share in hosts (ie 192.168.2.121 guest.home.network guest )
    > however this has not helped me at all.
    >
    > I can see the downstairs computer running 2007.1 but the two other
    > computers do no see each others shares.
    >
    > I would be grateful for any advise, although I think that not
    > networking is well appalling for mandy and the MCC app should work, as
    > in omg. As in ---???



    SOLUTION (for me anyway) !!!

    In services nfs-utils and nfs-common are not starting or are being
    stopped, I started both -but it didn't help. However when I stopped
    resolvconf and went back to MCC's "Use nfs shares" -there they were(!),
    immediately -no search -no clicking on the caret (arrow) to displays
    the host's shares, they just popped in and pip pop I mounted them.

    So here's my solution for you people looking through here to figure out
    where your shares went. Make sure nfs-utlis and nfs-common are
    running in (MCC-->System-->Services) and in the same window - stop
    resolvconf.

    You will get info as they trun on and off and on again.

    And to the people that will be fixing this (soon) this is sum useful
    info top start looking.

  19. Re: SOLVED!! NFS sharing not working 2008

    On Thu, 15 Nov 2007 14:02:14 -0400, Marlene wrote:

    > SOLUTION (for me anyway) !!!
    >
    > In services nfs-utils and nfs-common are not starting or are being
    > stopped, I started both -but it didn't help. However when I stopped
    > resolvconf and went back to MCC's "Use nfs shares" -there they were(!),
    > immediately -no search -no clicking on the caret (arrow) to displays
    > the host's shares, they just popped in and pip pop I mounted them.
    >
    > So here's my solution for you people looking through here to figure out
    > where your shares went. Make sure nfs-utlis and nfs-common are
    > running in (MCC-->System-->Services) and in the same window - stop
    > resolvconf.


    Hmmm, resolvconf rebuilds /etc/resolv.conf

    If just stopping that, that would indicate dns resolver could not get target
    node ip from /etc/hosts for share resolution, asked the dns server
    indicated in /etc/resolv.conf for the target ip.
    Usually that nameserver is the isp or modem's dns server.

    What happened after that causes the problem.

    That is why I have FQDNs with alias for all nodes on my lan in
    everyone's /etc/hosts file.

+ Reply to Thread