a) Description of problem:

Gluster commands such as 'gluster peer status' or 'gluster volume list' does not output anything. This happens only on a single node in a 5 node RHS3.0 cluster. 

b) Version-Release number of selected component (if applicable):

glusterfs-3.6.0.42-1.el6rhs.x86_64

c) How reproducible:

Always.

d) Steps to Reproduce:

Unfortunately, none.

e) Actual results:

Any gluster commands executed from a specific node in a 5 node RHS3.0 cluster doesn't return anything. This is not seen on the other nodes in the cluster. 

f) Expected results:

gluster commands should return proper outputs.

g) Additional info:

<snip>
[root@node01 ~]# service glusterd status
glusterd (pid  16392) is running...

[root@node01 ~]# gluster peer status
<<no output>>
[root@node01 ~]# gluster volume list
<<no output>>
[root@node1 ~]# tail -f /var/log/glusterfs/cli.log
[2015-02-26 10:35:02.171720] E [mem-pool.c:353:mem_get0] (-->/usr/sbin/gluster(cli_rpc_init+0x2c) [0x40982c] (-->/usr/lib64/libglusterfs.so.0(dict_new+0xb) [0x308381bd6b] (-->/usr/lib64/libglusterfs.so.0(get_new_dict_full+0x25) [0x308381bc25]))) 0-mem-pool: invalid argument
[2015-02-26 10:35:32.856346] E [mem-pool.c:353:mem_get0] (-->gluster(cli_rpc_init+0x2c) [0x40982c] (-->/usr/lib64/libglusterfs.so.0(dict_new+0xb) [0x308381bd6b] (-->/usr/lib64/libglusterfs.so.0(get_new_dict_full+0x25) [0x308381bc25]))) 0-mem-pool: invalid argument
</snip>

An strace on 'gluster volume status' shows that /var/log/glusterd/cli.log is opened but the normal details are not written to it. The only messages written are:

<snip>
[2015-03-10 15:54:27.550362] E [mem-pool.c:353:mem_get0] (-->gluster(cli_rpc_init+0x2c) [0x40982c] (-->/usr/lib64/libglusterfs.so.0(dict_new+0xb) [0x308381bd6b] (-->/usr/lib64/libglusterfs.so.0(get_new_dict_full+0x25) [0x308381bc25]))) 0-mem-pool: invalid argument
</snip>