[lustre-discuss] Problems moving an OSS from an old Lustre installation to a new one
massimo.sgaravatto at pd.infn.it
Mon Jul 27 22:46:32 PDT 2015
Indeed we forgot to shutdown the old MGS/MDS before "attaching" the old
OSS to the new one, but I wonder why this caused problems.
We switched the old MGS/MDS off as soon as we realized about those
messages, but we still saw problems after this operation.
Our idea now would be to retry the mkfs operations (this time they would
be done with the old MDS switched down)
Do you see some better options ?
On 28/07/2015 07:38, Oliver Mangold wrote:
> On 28.07.2015 07:17, Massimo Sgaravatto wrote:
>> Some interesting messages found in the syslog of the "moved" OSS:
>> Jul 24 14:56:25 t2-oss-03 kernel: Lustre: cmswork-OST0003: Received
>> MDS connection from 10.60.16.8 at tcp, removing former export from
>> 10.60.16.38 at tcp
>> Jul 24 14:56:27 t2-oss-03 kernel: Lustre: cmswork-OST0003: already
>> connected client cmswork-MDT0000-mdtlov_UUID \
>> (at 10.60.16.8 at tcp) with handle 0xdb376ec08bf7d020. Rejecting client
>> with the same UUID trying to reconnect with\
>> handle 0x6dffb49bb9b3bc70
>> 10.60.16.8 is the IP of the old MDS
>> 10.60.16.38 is the IP of the new MDS
>> For the the being we disabled the OSTs hosted on the "moved" OSS so
>> that new objects are not written there.
>> Any idea what the problem is and how we could recover the system ?
> Do I see it correctly, that the old MGS/MDS is still up and running? I
> understand it that way, that it still tries to find a OST at
> 10.60.16.9 at tcp (that info is stored in the llog on the MGS). But I'm
> confused also, why it should think that the new OST is the one it is
> looking for. It has a new UUID, so it should be detected. Anyway, I
> would first shutdown the old MGS/MDS before I tried to write any more
> data to the new OST.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 1877 bytes
Desc: S/MIME Cryptographic Signature
More information about the lustre-discuss