[Gluster-users] Geo-replication connection addresses

[Audrius Butkevicius]

Hi,

I am having trouble setting up geo replication, somehow the gsync client
decides to use a loopback address, for what I assume is a reverse
connection:

[2015-10-12 00:07:11.409406] I [monitor(monitor):362:distribute] <top>:
slave bricks: [{'host': 'other-peer.com', 'dir': '/data/website'}]
[2015-10-12 00:07:11.424458] I [monitor(monitor):383:distribute] <top>:
worker specs: [('/data/website',
'ssh://ubuntu@other-peer.com:gluster://localhost:website',
1)]
[2015-10-12 00:07:11.753465] I [monitor(monitor):221:monitor] Monitor:
------------------------------------------------------------
[2015-10-12 00:07:11.753704] I [monitor(monitor):222:monitor] Monitor:
starting gsyncd worker
[2015-10-12 00:07:11.977433] I [gsyncd(/data/website):649:main_i] <top>:
syncing: gluster://localhost:website -> ssh://ubuntu@other-peer.com:
gluster://localhost:website
[2015-10-12 00:07:11.990741] I [changelogagent(agent):75:__init__]
ChangelogAgent: Agent listining...
[2015-10-12 00:07:14.449852] E
[syncdutils(/data/website):252:log_raise_exception] <top>: connection to
peer is broken
[2015-10-12 00:07:14.455070] E [resource(/data/website):222:errlog] Popen:
command "ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i
/var/lib/glusterd/geo-replication/secret.pem -oControlMaster=auto -S
/tmp/gsyncd-aux-ssh-2O1R6B/b753bb9b7fed2b5f2777630d7feb9b12.sock
ubuntu at other-peer.com /usr/lib/x86_64-linux-gnu/glusterfs/gsyncd
--session-owner 052d5807-8bbf-4ce6-97b5-9d2f15e7a16e -N --listen --timeout
120 gluster://localhost:website" returned with 1, saying:
[2015-10-12 00:07:14.455183] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.275144] I [cli.c:721:main] 0-cli: Started running
/usr/sbin/gluster with version 3.7.5
[2015-10-12 00:07:14.455287] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.288762] I [cli.c:608:cli_rpc_init] 0-cli:
Connecting to remote glusterd at localhost
[2015-10-12 00:07:14.455384] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.482249] I [MSGID: 101190]
[event-epoll.c:632:event_dispatch_epoll_worker] 0-epoll: Started thread
with index 1
[2015-10-12 00:07:14.455472] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.482392] I [socket.c:2355:socket_event_handler]
0-transport: disconnecting now
[2015-10-12 00:07:14.455548] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.482987] I [cli-rpc-ops.c:6348:gf_cli_getwd_cbk]
0-cli: Received resp to getwd
[2015-10-12 00:07:14.456195] E [resource(/data/website):226:logerr] Popen:
ssh> [2015-10-12 00:07:16.483043] I [input.c:36:cli_batch] 0-: Exiting
with: 0
[2015-10-12 00:07:14.456406] I [syncdutils(/data/website):220:finalize]
<top>: exiting.
[2015-10-12 00:07:14.457257] I [repce(agent):92:service_loop] RepceServer:
terminating on reaching EOF.
[2015-10-12 00:07:14.457413] I [syncdutils(agent):220:finalize] <top>:
exiting.
[2015-10-12 00:07:15.458548] I [monitor(monitor):282:monitor] Monitor:
worker(/data/website) died in startup phase
[2015-10-12 00:07:25.785124] I [monitor(monitor):221:monitor] Monitor:
------------------------------------------------------------


I also had a few troubles around the fact that both machines are behind AWS
NAT on different regions, so they somehow advertise to each other their
hostnames, which resolve to the local VPN(LAN) address, but I worked around
that by adding entries in /etc/hosts.

Thanks,
Audrius.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20151012/545c4127/attachment.html>