[Bugs] [Bug 1162479] New: replace-brick doesn't work fine .

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1162479

            Bug ID: 1162479
           Summary: replace-brick doesn't work fine .
           Product: GlusterFS
           Version: 3.4.2
         Component: cli
          Severity: urgent
          Assignee: bugs at gluster.org
          Reporter: wangqy8848 at 163.com
                CC: bugs at gluster.org, gluster-bugs at redhat.com
   External Bug ID: CentOS 1
   External Bug ID: CentOS 1



Description of problem:
I build a replica 2 volume ,and take a replace-brick test. but it does not work
well.

Version-Release number of selected component (if applicable):
I prepared 3 virtual PCs with CentOS 6.5 and gluster 3.4.2.
And I also do the same test with gluster 3.5.2 and get the same result.

How reproducible:
I did this test several times，and always get same results.

Steps to Reproduce:
1.build a replVolume Name: testrep
Type: Replicate
Volume ID: 172e5c5c-fb94-4f8b-9ed4-b764b9c5d6cd
Status: Started
Number of Bricks: 1 x 2 = 2
Transport-type: tcp
Bricks:
Brick1: data-node3:/brick/testrep
Brick2: data-node4:/brick/testrep
ica volume, such as:

2.mount it on a client,and write something into it.

3.do the replace-brick oprations:
  start replace and trace its log.
  command like this:
    # gluster volume replace-brick testrep data-node4:/brick/testrep
data-node5:/brick/testrep --log-level=TRACE --log-file=/tmp/replacelog start

  and it will take a long time to return. and then the gluster crashed.

  after that, i try to get volume info. it will also take a long time ,and
return no volume. and get volume status ,it will return nothing.
  command like this:
    # gluster volume status
    (there is nothing return)
    # gluster volume info
    No volumes present

the replace-brick opration cann't continue.

Additional info:
try to get something from the trace log.

[2014-11-11 07:19:01.558436] T [rpc-clnt.c:424:rpc_clnt_reconnect] 0-glusterfs:
attempting reconnect
[2014-11-11 07:19:01.558543] T [socket.c:2675:socket_connect]
(-->/lib64/libpthread.so.0(+0x79d1) [0x7fa5640539d1]
(-->/usr/lib64/libglusterfs.so.0(gf_timer_proc+0xc8) [0x7fa565ccd368]
(-->/usr/lib64/libgfrpc.so.0(rpc_clnt_reconnect+0x116) [0x7fa5654182e6])))
0-glusterfs: connect () called on transport already connected
[2014-11-11 07:19:01.558553] T [rpc-clnt.c:424:rpc_clnt_reconnect] 0-glusterfs:
attempting reconnect
[2014-11-11 07:19:01.558574] T [socket.c:2683:socket_connect] 0-glusterfs:
connecting 0x1417480, state=0 gen=0 sock=-1
[2014-11-11 07:19:01.558595] W [dict.c:1055:data_to_str]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(+0x68ec)
[0x7fa5629ab8ec]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(socket_client_get_remote_sockaddr+0xad)
[0x7fa5629affcd]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(client_fill_address_family+0x200)
[0x7fa5629afe80]))) 0-dict: data is NULL
[2014-11-11 07:19:01.558609] W [dict.c:1055:data_to_str]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(+0x68ec)
[0x7fa5629ab8ec]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(socket_client_get_remote_sockaddr+0xad)
[0x7fa5629affcd]
(-->/usr/lib64/glusterfs/3.5.2/rpc-transport/socket.so(client_fill_address_family+0x20b)
[0x7fa5629afe8b]))) 0-dict: data is NULL
[2014-11-11 07:19:01.558615] E [name.c:147:client_fill_address_family]
0-glusterfs: transport.address-family not specified. Could not guess default
value from (remote-host:(null) or transport.unix.connect-path:(null)) options

it will try it every 3sec.

at last it exit with 110:

[2014-11-11 07:19:04.527694] D [cli-cmd.c:381:cli_cmd_submit] 0-cli: Returning
110
[2014-11-11 07:19:04.527754] D [cli-rpc-ops.c:3531:gf_cli_replace_brick] 0-cli:
Returning 110
[2014-11-11 07:19:04.527768] D
[cli-cmd-volume.c:1578:cli_cmd_volume_replace_brick_cbk] 0-cli: frame->local is
not NULL (0x7fa5540009c0)
[2014-11-11 07:19:04.527794] I [input.c:36:cli_batch] 0-: Exiting with: 110

i wander if something wrong with my enverment.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.