Re: [Gluster-users] AUX GID > 16

Anush Shetty Fri, 18 May 2012 07:10:54 -0700

Hi Ling,

Was not able to reproduce it locally. Do you have a script which canreproduce this issue.


Let me know.

Thanks,
Anush

On 05/17/2012 07:38 AM, Ling Ho wrote:

Hi Anush,

Did you have a chance to try to reproduce the error? Should I expectthis feature/fix be supported in 3.3?

Let me know if I am doing the right thing to test it, or if I shoulddownload a different tarball to try it.


Thanks,
...
ling

On 05/07/2012 11:56 AM, Ling Ho wrote:

Hi Anush,

This is how I see the error. I added myself into more than 16 posix
groups  and I start seeing the "Transport endpoint is not connected"
error, and could no longer access the gluster file system anymore. I
have no problem when using another user (tstopr) that belongs to less
than 16 posix groups. Both server and client machines are running
RHEL5.8, kernel version 2.6.18-308.1.1, and glusterfs 3.3.0beta3-1 .

Under 3.2.6-1, I could at least see the file system content as both
users. Of couse I have problem with group permissions because of the rpc
16 gid limit.

[root@pssrv105 x86_64]# mount -t glusterfs psanabackup02:backup/tmp/mnt

[root@pssrv105 x86_64]# ls /tmp/mnt
backup_ana01_scratch.01-17-2012  backup_ana12_scratch.03-29-2012
backup_ana02_scratch.01-17-2012  backup_ana12_scratch.03-30-2012
backup_ana11_scratch.03-22-2012  backup_ana12_scratch.03-31-2012
backup_ana11_scratch.03-29-2012  backup_ana12_scratch.04-02-2012
backup_ana11_scratch.03-30-2012  backup_ana12_scratch.04-03-2012
backup_ana11_scratch.03-31-2012  backup_ana12_scratch.04-05-2012
backup_ana11_scratch.04-10-2012  backup_ana12_scratch.04-06-2012
backup_ana11_scratch.04-11-2012  backup_ana12_scratch.04-12-2012
backup_ana11_scratch.05-04-2012  backup_ana12_scratch.04-16-2012
backup_ana11_scratch.current     backup_ana12_scratch.current
backup_ana12_scratch.03-19-2012  test
backup_ana12_scratch.03-26-2012
[root@pssrv105 x86_64]# su - ling
-bash-3.2$ cd /tmp/mnt
-bash-3.2$ ls -al
ls: .: Transport endpoint is not connected
-bash-3.2$
-bash-3.2$ id
uid=10730(ling) gid=1107(re)

groups=1107(re),2293(amo00409),2294(amo02809),2295(amo01109),2306(amo14410),2307(amo10610),2308(amo10810),2309(amo15010),2310(amo10510),2341(ps-ioc),2352(amo19810),2353(amo22210),2392(ps-cxi),2406(amom0111),2475(cxi49812),10000(ps-users),10001(ps-pcds),10002(ps-icsmec)


-bash-3.2$ logout
[root@pssrv105 x86_64]# su - tstopr
pssrv105:~>  cd /tmp/mnt
pssrv105:mnt>  ls
backup_ana01_scratch.01-17-2012  backup_ana12_scratch.03-29-2012
backup_ana02_scratch.01-17-2012  backup_ana12_scratch.03-30-2012
backup_ana11_scratch.03-22-2012  backup_ana12_scratch.03-31-2012
backup_ana11_scratch.03-29-2012  backup_ana12_scratch.04-02-2012
backup_ana11_scratch.03-30-2012  backup_ana12_scratch.04-03-2012
backup_ana11_scratch.03-31-2012  backup_ana12_scratch.04-05-2012
backup_ana11_scratch.04-10-2012  backup_ana12_scratch.04-06-2012
backup_ana11_scratch.04-11-2012  backup_ana12_scratch.04-12-2012
backup_ana11_scratch.05-04-2012  backup_ana12_scratch.04-16-2012
backup_ana11_scratch.current     backup_ana12_scratch.current
backup_ana12_scratch.03-19-2012  test
backup_ana12_scratch.03-26-2012
pssrv105:mnt>  id
uid=10858(tstopr) gid=1109(xs) groups=1109(xs)


Thanks,

...
ling

On 05/06/2012 11:10 PM, Anush Shetty wrote:

Hi Ling,

Can you please give me the steps used in testing the issue so that I
could reproduce it locally.

-
Anush


On 05/05/2012 06:24 AM, Ling Ho wrote:

Hi,

I tried again, with clean install both on the server and client, and
recreated the volume. Both have these four rpms installed.

glusterfs-3.3.0beta3-1
glusterfs-server-3.3.0beta3-1
glusterfs-rdma-3.3.0beta3-1
glusterfs-fuse-3.3.0beta3-1

I tried both rdma and tcp protocol, and both gave the same results.

Thanks,
...
ling


On 05/02/2012 10:34 PM, Anush Shetty wrote:

Hi,

Can you please give it another try with a clean install? Also please
make sure that both server and clients are of the same version.

-
Anush

On 05/03/2012 07:20 AM, Ling Ho wrote:

Hello,

I am looking at the changes done for this bug 767229.

I downloaded and installed 3.3.0beta3-1 which seems to have included

the changes. But when I make myself a member of more than 16groups, I

can't access any directory nor files under my mounted glusterfs file
system.  I got a "Transport endpoint is not connected" error. I have
no problem accessing the directories as root user, or as a different
user who belongs to less than 16 groups.

This is what I get in the log on the client machine:

[2012-05-02 18:48:11.121567] W [xdr-rpcclnt.c:88:rpc_request_to_xdr]
0-rpc: failed to encode call msg
[2012-05-02 18:48:11.121600] E
[rpc-clnt.c:1268:rpc_clnt_record_build_record] 0-backup-client-0:
Failed to build record header
[2012-05-02 18:48:11.121615] W [rpc-clnt.c:1328:rpc_clnt_record]
0-backup-client-0: cannot build rpc-record
[2012-05-02 18:48:11.121628] W [rpc-clnt.c:1467:rpc_clnt_submit]
0-backup-client-0: cannot build rpc-record
[2012-05-02 18:48:11.121641] W
[client3_1-fops.c:2528:client3_1_opendir_cbk] 0-backup-client-0: