[Gluster-users] auth.allow behavior?

[Harry Mangalam]

On Tuesday 01 November 2011 16:53:56 Joe Landman wrote:
 On 11/01/2011 06:09 PM, Harry Mangalam wrote:

> > What else sets the authentication / permission correctly?
> 
> gluster volume set g6 auth.allow 192.168.*,128.*

That doesn't cause an error and it does cause itself to be inserted 
into the volume info:
==================================
root at pbs3:~
561 $ gluster volume info

Volume Name: g6
Type: Distribute
Status: Stopped
Number of Bricks: 6
Transport-type: tcp
Bricks:
Brick1: pbs1:/data2
Brick2: pbs2:/data2
Brick3: pbs3:/data2
Brick4: pbs3:/data
Brick5: dabrick:/data2
Brick6: hef:/data2
Options Reconfigured:
auth.allow: 128.200.*,128.195.*,10.255.89.*,10.255.78.*
==================================

AND after volume stop/start (but NOT before), it DOES come back for 
the clients that could originally mount it.

AND IT NOW WORKS FOR THE OTHER clients as well.


So: 
- auth.allow supports multiple addresses IF they're separated with ',' 
but not with spaces (not documented, more silent failures)

- volumes have to be stopped and restarted to propagate these changes 
(another silent .. if not failure, the absence of a warning that 
indicates that on auth.allow (and all other option?) changes, you have 
to restart the volume to activate it.

This does not seem to be documented in the Gluster_FS_3.2_Admin_Guide 
that I have.  I guess I should be going thru the wiki on this...

Joe gets an ebeer (12pack!) :)

Thanks Joe!

hjm




-- 
Harry Mangalam - Research Computing, OIT, Rm 225 MSTB, UC Irvine
[ZOT 2225] / 92697  Google Voice Multiplexer: (949) 478-4487 
MSTB Lat/Long: (33.642025,-117.844414) (paste into Google Maps)
--
This signature has been OCCUPIED!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20111101/344b40e9/attachment.html>