Fwd: Error in Setup File Server Cluster with Samba

GiangCoi Mr ltrgiang86 at gmail.com
Tue May 30 06:17:47 UTC 2017


Hi Team.
At the moment, I am installing File Server Samba Cluster as follow diagram



File Server 01 and 02 connect to SAN 01 and SAN 02 through iSCSI. In both
File Server I install and configure GlusterFS to share folder /data for
shared files to everyone.

​File configure ctdb as follow

vi /data/lock/ctdb

CTDB_RECOVERY_LOCK=/data/lock/lockfile
#CIFS only
CTDB_PUBLIC_ADDRESSES=/etc/ctdb/public_addresses
CTDB_MANAGES_SAMBA=yes
#CIFS only
CTDB_NODES=/etc/ctdb/nodes


File public_address
vi /data/lock/public_addresses
10.1.1.202/24 eth0

File nodes
vi /data/lock/nodes
172.16.0.1
172.16.0.2


File smb.conf

cp /etc/samba/smb.conf /data/lock/smb.conf

vi /data/lock/smb.conf

clustering = yes
idmap backend = tdb2
private dir = /data/lock



[share]

comment = Gluster and CTDB based share
path = /data/share
read only = no
writable = yes
valid users = jon

I create soft link
ln -s /data/lock/ctdb /etc/sysconfig/ctdb

ln -s /data/lock/nodes /etc/ctdb/nodes
ln -s /data/lock/public_addresses /etc/ctdb/public_addresses

ln -s /data/lock/smb.conf /etc/samba/


Everything I setup, It's ok. When I test network down with 2 Client, I
disconnect eth0 in File Server 01, then

- Client 01: 10.1.15.200 (in other subnet) is copying data to File Server
01 interrupted and the Client 01 have lost connection to File Server. I saw
log from log.ctdb: "sendto failed, don't send tickle ACK to IP 10.1.15.200"

 - Client 03: 10.1.1.210 (in same subnet) is copying data to File Server 01
normally.


I am using Cisco Layer 3 to routing inter VLAN.


So, how I do to fix this issue?


Regards,

Giang
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Diagram.jpg
Type: image/jpeg
Size: 36346 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20170530/72fcd1e4/Diagram.jpg>


More information about the samba-technical mailing list