[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Linux-cluster] GFS and samba problem in Fedora, again



Hi Abhi,

I re-compiled ccsd with debug option, and I attach the logs of each node. CMAN works ok and quored is regained: Nov 10 13:30:54 tigreton kernel: CMAN: Waiting to join or form a Linux-cluster
Nov 10 13:31:24 tigreton kernel: CMAN: sending membership request
Nov 10 13:31:24 tigreton kernel: CMAN: got node nocilla.fib.upc.es
Nov 10 13:31:24 tigreton kernel: CMAN: quorum regained, resuming activity
Nov 10 13:36:45 tigreton kernel: CMAN: we are leaving the cluster. Removed


daemon.log:
...........
Nov 10 13:30:44 tigreton ccsd[3951]: [ccsd.c:87] Unable to create IPv6 socket:: Address family not supported by protocol ---> says that because I disabled IPV6 (the result with it was the same).
Nov 10 13:30:44 tigreton ccsd[3951]: [ccsd.c:99] Using IPv4
Nov 10 13:30:44 tigreton ccsd[3951]: [ccsd.c:878] Entering setup_local_socket
...........

In Red Hat documentation there is a daemon called clvmd. We didn't install because it wasn't in rpms, and seems unuseful because two nodes sees LVM logical partitions without it. In my test with RHES I didn't install it also, and GFS+samba worked, but I asked if it could be a relation between clvmd and locks problems with samba? Do you think that is important enough to try install it?

We are planning to buy RHES4 because, in test environment GFS and samba works better, and we're interested in support. Is GFS included in RHES4 for academic use, or I have to download it from SVN? What kind of support RH gives if we buy RHES?
Thanks a lot,

Sandra Hernández

Abhijith Das wrote:

Hi Sandra,
I'm assuming you're talking about the RHEL4 branch of CVS. I'm not 100% sure that the cluster-suite and GFS are supported for FC5. (Somebody on this list can confirm). The error however looks like ccsd is unable to talk to cman. At what point do you see this error? Are you using the init-scripts? I'd suggest that you compile ccsd with the DEBUG flag and try starting all the components by hand. Hopefully that'll give us more information.

Thanks,
--Abhi

Attachment: daemonNODE1.log.gz
Description: GNU Zip compressed data

Attachment: daemonNODE2.log.gz
Description: GNU Zip compressed data


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]