[Gluster-users] glusterfs with samba ctdb

[Daniel Filipazzi]

I was about to set upp glusterfs with samba ctdb, i got just about every thing to work up untill i was about to access the samba share .
from a linux client i can see the share but i get "unable to mount location"
and on a windows client "gluser-share is not accessible"


Initially i thought that this was an samba / permissions issue but no, ive added users and set permissions to 777.
Another clue here is what i get from the log.

-----------------------------------------------------------------------------
Feb  2 14:36:24 centosgfs1 smbd[25962]: [2016/02/02 14:36:24.985817,  0] lib/ctdbd_conn.c:624(ctdb_handle_message)
Feb  2 14:36:24 centosgfs1 smbd[25962]:   Got cluster reconfigure message
Feb  2 14:36:44 centosgfs1 smbd[25962]: [2016/02/02 14:36:44.989928,  0] lib/ctdbd_conn.c:624(ctdb_handle_message)
Feb  2 14:36:44 centosgfs1 smbd[25962]:   Got cluster reconfigure message
Feb  2 14:37:22 centosgfs1 ctdbd: CTDB is being run without syslog enabled.  Logs will be in /var/log/log.ctdb
Feb  2 14:37:48 centosgfs1 root: All required GPFS resources are available. CNFS part is healthy.
Feb  2 14:44:18 centosgfs1 smbd[15061]: [2016/02/02 14:44:18.662763,  0] modules/vfs_glusterfs.c:264(vfs_gluster_connect)
Feb  2 14:44:18 centosgfs1 smbd[15061]:   share: Failed to initialize volume (Transport endpoint is not connected)
Feb  2 14:44:19 centosgfs1 smbd[15061]: [2016/02/02 14:44:19.649592,  0] smbd/service.c:941(make_connection_snum)
Feb  2 14:44:19 centosgfs1 smbd[15061]:   make_connection: VFS make connection failed!
Feb  2 15:37:46 centosgfs1 smbd[11652]: [2016/02/02 15:37:46.260461,  0] lib/ctdbd_conn.c:624(ctdb_handle_message)
Feb  2 15:37:46 centosgfs1 smbd[11652]:   Got cluster reconfigure message
Feb  2 15:37:53 centosgfs1 smbd[15052]: [2016/02/02 15:37:53.755840,  0] lib/ctdbd_conn.c:681(ctdbd_socket_handler)
Feb  2 15:37:53 centosgfs1 smbd[11661]: [2016/02/02 15:37:53.755932,  0] lib/ctdbd_conn.c:681(ctdbd_socket_handler)
Feb  2 15:37:53 centosgfs1 smbd[11661]:   packet_fd_read failed: NT_STATUS_END_OF_FILE
Feb  2 15:37:53 centosgfs1 smbd[11661]: [2016/02/02 15:37:53.758719,  0] lib/ctdbd_conn.c:57(cluster_fatal)
Feb  2 15:37:53 centosgfs1 smbd[11661]:   cluster fatal event: ctdbd died
Feb  2 15:37:53 centosgfs1 smbd[11661]:    - exiting immediately
Feb  2 15:37:53 centosgfs1 smbd[15052]:   packet_fd_read failed: NT_STATUS_END_OF_FILE
Feb  2 15:37:53 centosgfs1 smbd[15052]: [2016/02/02 15:37:53.758867,  0] lib/ctdbd_conn.c:57(cluster_fatal)
Feb  2 15:37:53 centosgfs1 smbd[15052]:   cluster fatal event: ctdbd died
Feb  2 15:37:53 centosgfs1 smbd[15052]:    - exiting immediately
Feb  2 15:37:54 centosgfs1 ctdbd: CTDB is being run without syslog enabled.  Logs will be in /var/log/log.ctdb
Feb  2 15:38:14 centosgfs1 root: All required GPFS resources are available. CNFS part is healthy.
Feb  2 15:38:25 centosgfs1 smbd[9796]: [2016/02/02 15:38:25.198307,  0] lib/ctdbd_conn.c:624(ctdb_handle_message)
Feb  2 15:38:25 centosgfs1 smbd[9796]:   Got cluster reconfigure message
Feb  2 15:38:36 centosgfs1 smbd[10492]: [2016/02/02 15:38:36.774688,  0] modules/vfs_glusterfs.c:264(vfs_gluster_connect)
Feb  2 15:38:36 centosgfs1 smbd[10492]:   share: Failed to initialize volume (Transport endpoint is not connected)
Feb  2 15:38:37 centosgfs1 smbd[10492]: [2016/02/02 15:38:37.755447,  0] smbd/service.c:941(make_connection_snum)
Feb  2 15:38:37 centosgfs1 smbd[10492]:   make_connection: VFS make connection failed!
Feb  2 15:38:37 centosgfs1 smbd[10498]: [2016/02/02 15:38:37.966779,  0] modules/vfs_glusterfs.c:264(vfs_gluster_connect)
Feb  2 15:38:37 centosgfs1 smbd[10498]:   share: Failed to initialize volume (Transport endpoint is not connected)
Feb  2 15:38:38 centosgfs1 smbd[10498]: [2016/02/02 15:38:38.952001,  0] smbd/service.c:941(make_connection_snum)
Feb  2 15:38:38 centosgfs1 smbd[10498]:   make_connection: VFS make connection failed!


-----------------------------------------------------------------------------------------------------------------------------------

I have a feeling that this is a pretty common issue, question is whats the fix and is the actual issue here?


More over since glusterfs automaticly generates configuration in smb.conf i gather that nothing should be modified, though the parameter: path = / leaves me a bit puzzled. Im guessing that the parameter: glusterfs:volume = share is the one that does the trick.



Best regards
Daniel Filipazzi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160203/a49663a2/attachment.html>