Gluster users - Jan 2010 - upgrade path 2.0.0rc1 -> 2.0.9 ?

Hello all,

I have a small gluster setup with two servers and two clients doing 
client-side replication to the servers.  All of the nodes are currently 
running 2.0.0rc1 .

In preparation for an upgrade to 2.0.9 i set up a series of virtual 
machines with the same OS and environment as the actual machines.  After 
compiling and installing the new binaries, i restarted the various 
services and mounts, and everything hummed along nicely.

Unfortunately, this was not the case for my actual machines. :(  I 
upgraded one of the servers, then restarted the gluster service, and saw 
nothing but this over and over in the logs (at high speed) :

[2010-01-26 11:09:26] N [server-protocol.c:7825:notify] server: 
10.0.0.16:853 disconnected

10.0.0.16 being, in this case, one of the clients.  On the same client, 
the log was full of these messages :

2010-01-26 11:18:13 E [client-protocol.c:5604:client_setvolume_cbk] 
nasdash-01: SETVOLUME on remote-host failed: UUID not specified

nasdash-01 being, in this case, the server noted above.

I thus have two questions :

1. What does the ? SETVOLUME on remote-host failed: UUID not specified ? 
error actually mean ?  I realise that it means i need to specify the 
UUID - what is that, why do i need to specify it, and how do i go about 
that ?  (A search for ? UUID ? on the gluster wiki produces zero hits.)

2. What is the Official Gluster Approved Upgrade Methodology for going 
from 2.0.0rc1 to 2.0.9 ?

Thank you.

(p.s. We're not ready to move to the 3.0.x branch...)


-- 
Daniel Maher <dma+gluster AT witbe DOT net>

Hi Daniel,

Have you upgraded all the clients and servers before starting glusterfs? If
not, you need to upgrade all the servers and clients before restarting
glusterfs.

regards,
On Tue, Jan 26, 2010 at 3:31 PM, Daniel Maher
<dma+gluster at witbe.net<dma%2Bgluster at
witbe.net>
> wrote:
> Hello all,
>
> I have a small gluster setup with two servers and two clients doing
> client-side replication to the servers.  All of the nodes are currently
> running 2.0.0rc1 .
>
> In preparation for an upgrade to 2.0.9 i set up a series of virtual
> machines with the same OS and environment as the actual machines.  After
> compiling and installing the new binaries, i restarted the various services
> and mounts, and everything hummed along nicely.
>
> Unfortunately, this was not the case for my actual machines. :(  I upgraded
> one of the servers, then restarted the gluster service, and saw nothing but
> this over and over in the logs (at high speed) :
>
> [2010-01-26 11:09:26] N [server-protocol.c:7825:notify] server:
> 10.0.0.16:853 disconnected
>
> 10.0.0.16 being, in this case, one of the clients.  On the same client, the
> log was full of these messages :
>
> 2010-01-26 11:18:13 E [client-protocol.c:5604:client_setvolume_cbk]
> nasdash-01: SETVOLUME on remote-host failed: UUID not specified
>
> nasdash-01 being, in this case, the server noted above.
>
> I thus have two questions :
>
> 1. What does the ? SETVOLUME on remote-host failed: UUID not specified ?
> error actually mean ?  I realise that it means i need to specify the UUID -
> what is that, why do i need to specify it, and how do i go about that ?  (A
> search for ? UUID ? on the gluster wiki produces zero hits.)
>
> 2. What is the Official Gluster Approved Upgrade Methodology for going from
> 2.0.0rc1 to 2.0.9 ?
>
> Thank you.
>
> (p.s. We're not ready to move to the 3.0.x branch...)
>
>
> --
> Daniel Maher <dma+gluster AT witbe DOT net>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>


-- 
Raghavendra G