Problems with OVH NFS NAS

[egrueda]

Hi all. I wanted to make some testings placing a CT inside a NFS share from the OVH's NAS service.
I added a new storage ponting to that NAS, it worked right and I made som backup tests, eveything went fine.
Then I tried to move a CT to the NAS storage, but couldn't start ir properly, so went back.

I wanted to make a clean test. I created a new CT directly in the NAS storage (remmember, nfs)
Creation of the container had no problems (a bit slower uncompressing, of course)
Then I started this CT and had exactly the same problem as before.
CT is started and mounted, but it looks like it's not booting.
CT is 99% CPU, I can connect to console but it seems hanged on boot.
It won't listen for a local "init 0" (not found) and took a very long time (10 mins) and 3 attempts to stop it using vzctl

The only log I see is about the NAS server (ip 11.22.33.44):

kernel: ct0 lockd: server 11.22.33.44 not responding, still trying
kernel: ct0 lockd: server 11.22.33.44 OK

When CT is stopped, those messages disapear from syslog (debian server)

I don't understand why the machine is not booting?
Anyone can help me? Thanks.

 

[egrueda]

This is the startup log:

# vzctl --verbose start 102Starting container ...
Container is mounted
Adding IP address(es): 192.168.0.102
Running container script: /etc/vz/dists/scripts/redhat-add_ip.sh
/bin/cp: preserving permissions for `/etc/sysconfig/network.4': Operation not supported
/bin/cp: preserving ACL for `/etc/sysconfig/network.4': Operation not supported
 ERROR: Can't copy file /etc/sysconfig/network
Setting CPU units: 1000
Setting CPUs: 1
Running container script: /etc/vz/dists/scripts/redhat-set_hostname.sh
/bin/cp: preserving permissions for `/etc/hosts.9': Operation not supported
/bin/cp: preserving ACL for `/etc/hosts.9': Operation not supported
 ERROR: Can't copy file /etc/hosts
Configure veth devices: veth102.0 
Adding interface veth102.0 to bridge vmbr0 on CT0 for CT102
Container start in progress...

and

# vzctl --verbose status 102
CTID 102 exist mounted running

And there it goes again:

# vzctl exec 102 ps ax  PID TTY      STAT   TIME COMMAND
    1 ?        Ds     0:00 init boot     
    2 ?        S      0:00 [kthreadd/102]
    3 ?        S      0:00 [khelper/102]
   11 ?        R      3:28 [init-logger]
   60 ?        S<s    0:00 /sbin/udevd -d
  201 ?        Rs     0:00 ps ax

I suppose permissions and ACL errors are due to the nfs server

 

[egrueda]

After more than 30 minutes "starting" the CT, I see this processes running on it:

# vzctl exec 102 ps ax                                 PID TTY      STAT   TIME COMMAND
    1 ?        Ds     0:00 init [3]      
    2 ?        S      0:00 [kthreadd/102]
    3 ?        S      0:00 [khelper/102]
   11 ?        R     39:11 [init-logger]
   60 ?        S<s    0:00 /sbin/udevd -d
  377 ?        Ss     0:00 /bin/bash /etc/rc.d/rc 3
  485 ?        Z      0:00 [ifup-eth] <defunct>
  544 ?        Z      0:00 [ifup-eth] <defunct>
  575 ?        S      0:00 /bin/bash /etc/rc3.d/S12syslog start
  578 ?        S      0:00 /bin/bash -c ulimit -S -c 0 >/dev/null 2>&1 ; syslogd -m 0
  579 ?        S      0:00 syslogd -m 0
  580 ?        Ds     0:00 syslogd -m 0
  594 ?        Rs     0:00 ps ax

It looks like it's trying to get up, and keeps going on:

# vzctl exec 102 ps ax        PID TTY      STAT   TIME COMMAND
    1 ?        Ss     0:00 init [3]      
    2 ?        S      0:00 [kthreadd/102]
    3 ?        S      0:00 [khelper/102]
   11 ?        R     43:36 [init-logger]
   60 ?        S<s    0:00 /sbin/udevd -d
  377 ?        Ss     0:00 /bin/bash /etc/rc.d/rc 3
  580 ?        Ss     0:00 syslogd -m 0
  597 ?        S      0:00 /bin/bash /etc/rc3.d/S55sshd start
  600 ?        S      0:00 /bin/bash /etc/rc3.d/S55sshd start
  601 ?        S      0:00 /bin/bash /etc/rc3.d/S55sshd start
  602 ?        D      0:00 runlevel
  735 ?        Rs     0:00 ps ax

 

[mir]

How is your NFS share exported?
I have seen strange things happening if the share is not exported with the option: no_root_squash

 

[egrueda]

That's the problem: I have no idea about it, I just have a control panel to authorize IPs but NFS NAS is out of my control
:-(

 

[mir]

Are you not able to login to the NAS box via a secure shell?
If you have this option it will be easy to verify: sudo exportfs -v
will show mount options.

 

[egrueda]

Mir, NAS is managed, I can't put my hands on it, it's external, OVH just let me see my partition, so I cannot log into the NAS neither see its configuration.
That's why I need to guest what is happening to my CT