[Gluster-devel] Multiple independent glusterfs daemons on the same host? Using vservers?

[Martin Fick]

Is it supposed to be possible to run multiple
independent glusterfsd daemons on the same host but
listening to different ports or ips?  I tried it with
different ports and the second instance of the daemon
refused to start saying that it was already running. 
I tried this manually (not with startup scripts) so it
should not have been checking a PID file.

Since I am a little adventurous, I was not willing to
give up so easily.  I am running debian kernel
2.6.17-1-vserver-686 which, obviously has vserver
support compiled in.  So I figured why not try running
the daemon inside a vserver.  Surprise, it ran, at the
same time as a daemon on the host!  Both instances
seemed to not interfere with each other.  I even used
the same volumes name to see if I could confuse it, it
seemed undaunted.  So the question is, what makes this
different from running two instances on different
ports?  What kind of check is being done in glusterfsd
and what are the reasons that this should be
prevented?

Lastly, while I was able to make things work withing a
vserver, I was only able to mount the client
filesystem inside the same vserver from which it was
exported, not on the host.  On the host when trying to
mount the fs exported from within a guest I get:

2008-03-05 13:28:53 E
[protocol.c:253:gf_block_unserialize_transport]
client: EOF from peer (192.168.1.75:9996)
2008-03-05 13:28:53 C [tcp.c:81:tcp_disconnect]
client: connection disconnected

Any ideas why this may be?

-Martin



      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping