[Bugs] [Bug 1149863] New: Option transport.socket.bind-address ignored
bugzilla at redhat.com
bugzilla at redhat.com
Mon Oct 6 19:13:45 UTC 2014
https://bugzilla.redhat.com/show_bug.cgi?id=1149863
Bug ID: 1149863
Summary: Option transport.socket.bind-address ignored
Product: GlusterFS
Version: mainline
Component: glusterd
Keywords: Triaged
Severity: urgent
Assignee: ndevos at redhat.com
Reporter: ndevos at redhat.com
CC: bugs at gluster.org
Blocks: 1149857
+++ This bug was initially created as a clone of Bug #1149857 +++
+++ +++
+++ This bug has been filed to get a fix in mainline, when +++
+++ a patch has been merged, please post it to the bugs +++
+++ listed in the 'blocks' field. +++
Description of problem: when setting transport.socket.bind-address it is
ignored except for actually binding the glusterd daemon. Basically NFS, SHD and
everything else tries to connect to 127.0.0.1:24007 instead of the actual
address bound. This means FUSE is the only option for connects and of course
the volume get corrupted shortly after.
Look in xlators/mgmt/glusterd/src/glusterd-utils.c:glusterd_nodesvc_start() for
possible fix. ie:
glusterd_nodesvc_start() calls runner_add_args( ... , "-s", "localhost", ...)
above provided by ndevos in IRC
Version-Release number of selected component (if applicable): 3.5.2
How reproducible: set transport.socket.bind-address option and restart. viola!
Steps to Reproduce:
1. set transport.socket.bind-address option
2. restart glusterfs-server
3.
Actual results: no NFS, no SHD and anything else that attempts to connect to
127.0.0.1:24007
Expected results: it should work
Additional info:
--- Additional comment from Charles Williams on 2014-10-06 21:04:47 CEST ---
forgot to add:
in the "volume management" section of glusterd.vol is where you can set this
"option transport.socket.bind-address 123.123.123.123"
Referenced Bugs:
https://bugzilla.redhat.com/show_bug.cgi?id=1149857
[Bug 1149857] Option transport.socket.bind-address ignored
--
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=jZfl2mFhfX&a=cc_unsubscribe
More information about the Bugs
mailing list