[lustre-discuss] strange problem mounting lustre

[Riccardo Veraldi]

Hello,
Actually I have a problem which could be similar to this:

https://groups.google.com/forum/#!topic/lustre-discuss-list/ED6rxVGuKM8

I am running oss and mds on Centos 6.4 with lustre 2.4

Suddently my JBOD array failed. It has something like 12 OST on it.
When I established it back,  one of the OST was not mounting anymore 
from new clients:

When I try to mount from new clients I get this error: failed: Function 
not implemented

But the old clients which has the OST already mounted are working fine. 
Anyway if I dismount the lustre partiion, they cannot mount it anymore 
with same error:

failed: Function not implemented

the mds is running not on the failed JBOD array but on another storage.

So I have no clue why this is happening. On the MDS and OSS side I have 
no error and no hint in the lustre logs when clients try to mount.
I have the error on client side and that particular OST cannot be 
mounted anymore.

I already did a fsck on the /mgsgdt device partition which is a LVM.
but this did not help fixing the problem.

How could I debug what is going wrong ?
It never happened before.
Looks like something is corrupted somewhere but clients which are 
already mounting the the OST (thru mds) are working like a charm.
Remounting after dismounting won't work.
I did not reboot the oss yet because I am afraid of the consequences.
So I first wrote here to have some hint.

Cheers

Rick

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20160513/fbf92ff6/attachment.htm>