Bug 763917 (GLUSTER-2185) - Add-brick on gluster cli starts gNFS even when the volume is not started yet
Summary: Add-brick on gluster cli starts gNFS even when the volume is not started yet
Keywords:
Status: CLOSED WORKSFORME
Alias: GLUSTER-2185
Product: GlusterFS
Classification: Community
Component: core
Version: 3.1.1
Hardware: All
OS: Linux
low
low
Target Milestone: ---
Assignee: Pranith Kumar K
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2010-12-02 20:46 UTC by Harshavardhana
Modified: 2015-03-23 01:04 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed:
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Embargoed:


Attachments (Terms of Use)

Description Harshavardhana 2010-12-02 17:46:22 UTC
same is the case with remove-brick

Comment 1 Harshavardhana 2010-12-02 20:46:11 UTC
This is unwise why would nfs/server start when its server/client process are not even started. 

This results in errors on the client side with gNFS server up.

Comment 2 Pranith Kumar K 2010-12-03 01:08:03 UTC
(In reply to comment #1)
> same is the case with remove-brick

Harsha,
   Could you please elaborate on what exactly you are observing. Please include the steps you used to get to this issue.

Pranith

Comment 3 Harshavardhana 2010-12-05 03:51:13 UTC
> Harsha,
>    Could you please elaborate on what exactly you are observing. Please include
> the steps you used to get to this issue.
> 
> Pranith


Simple procedure create a volume 

then add a new brick to the created volume, this automatically starts "gNFS" process, which is unwise and not necessary until the volume is started. 

This leads to lot of problems, when one is trying to change the topology over a stopped volume.

Comment 4 Pranith Kumar K 2010-12-13 08:47:54 UTC
(In reply to comment #3)
> > Harsha,
> >    Could you please elaborate on what exactly you are observing. Please include
> > the steps you used to get to this issue.
> > 
> > Pranith
> 
> 
> Simple procedure create a volume 
> 
> then add a new brick to the created volume, this automatically starts "gNFS"
> process, which is unwise and not necessary until the volume is started. 
> 
> This leads to lot of problems, when one is trying to change the topology over a
> stopped volume.

Hi Harsha,
   Here is the output of the steps you mentioned to reproduce the issue. It does not start the gNfs process. Please let me know if I am missing something.

pranith @ /etc/glusterd
17:12:09 :) $ sudo gluster volume create pranith `hostname`:/tmp/1
Creation of volume pranith has been successful. Please start the volume to access data.

pranith @ /etc/glusterd
17:12:26 :) $ ps aux | grep gluster
root      8152  0.1  0.3  48480  7512 pts/0    Sl+  17:12   0:00 glusterd --debug
pranith   8179  0.0  0.0   7624   912 pts/5    S+   17:12   0:00 grep --color=auto gluster
pranith  21115  0.0  0.5  35792 11824 ?        Ss   11:26   0:07 SCREEN -S gluster

pranith @ /etc/glusterd
17:12:38 :) $ sudo gluster volume add-brick pranith `hostname`:/tmp/2
Add Brick successful

pranith @ /etc/glusterd
17:12:53 :) $ ps aux | grep gluster
root      8152  0.0  0.3  48480  7548 pts/0    Sl+  17:12   0:00 glusterd --debug
pranith   8198  0.0  0.0   7624   912 pts/5    S+   17:12   0:00 grep --color=auto gluster
pranith  21115  0.0  0.5  35792 11824 ?        Ss   11:26   0:07 SCREEN -S gluster


The process started only after volume start

pranith @ /etc/glusterd
17:12:55 :) $ sudo gluster volume start pranith
Starting volume pranith has been successful

pranith @ /etc/glusterd
17:15:41 :) $ ps aux | grep gluster
root      8152  0.1  0.4  50020  9128 pts/0    Sl+  17:12   0:00 glusterd --debug
root      8208  0.5  0.5  76924 10132 ?        Ssl  17:15   0:00 /usr/local/sbin/glusterfsd --xlator-option pranith-server.listen-port=24009 -s localhost --volfile-id pranith.pranith-laptop.tmp-1 -p /etc/glusterd/vols/pranith/run/pranith-laptop-tmp-1.pid --brick-name /tmp/1 --brick-port 24009 -l /usr/local/var/log/glusterfs/bricks/tmp-1.log
root      8212  1.0  0.5  76924 10128 ?        Ssl  17:15   0:00 /usr/local/sbin/glusterfsd --xlator-option pranith-server.listen-port=24010 -s localhost --volfile-id pranith.pranith-laptop.tmp-2 -p /etc/glusterd/vols/pranith/run/pranith-laptop-tmp-2.pid --brick-name /tmp/2 --brick-port 24010 -l /usr/local/var/log/glusterfs/bricks/tmp-2.log
root      8218  4.0  1.8  99032 37644 ?        Ssl  17:15   0:00 /usr/local/sbin/glusterfs -f /etc/glusterd/nfs/nfs-server.vol -p /etc/glusterd/nfs/run/nfs.pid -l /usr/local/var/log/glusterfs/nfs.log
pranith   8227  0.0  0.0   7624   912 pts/5    S+   17:15   0:00 grep --color=auto gluster
pranith  21115  0.0  0.5  35792 11824 ?        Ss   11:26   0:07 SCREEN -S gluster

pranith @ /etc/glusterd
17:15:43 :) $ sudo gluster volume stop pranith
Stopping volume will make its data inaccessible. Do you want to Continue? (y/n) y
^[[A^[[AStopping volume pranith has been successful

pranith @ /etc/glusterd
17:15:52 :) $ ps aux | grep gluster
root      8152  0.1  0.4  50020  9132 pts/0    Sl+  17:12   0:00 glusterd --debug
pranith   8234  0.0  0.0   7624   908 pts/5    S+   17:15   0:00 grep --color=auto gluster
pranith  21115  0.0  0.5  35792 11824 ?        Ss   11:26   0:07 SCREEN -S gluster

Comment 5 Harshavardhana 2010-12-14 18:50:20 UTC
> Hi Harsha,
>    Here is the output of the steps you mentioned to reproduce the issue. It
> does not start the gNfs process. Please let me know if I am missing something.
> 
> pranith @ /etc/glusterd
> 17:12:09 :) $ sudo gluster volume create pranith `hostname`:/tmp/1
> Creation of volume pranith has been successful. Please start the volume to
> access data.
> 
You might want start volume here after creation and then stop it.  Then try add-brick again, even if that didn't start nfs server you can close the bug as Invalid.

Comment 6 Pranith Kumar K 2010-12-15 02:46:21 UTC
(In reply to comment #5)
> > Hi Harsha,
> >    Here is the output of the steps you mentioned to reproduce the issue. It
> > does not start the gNfs process. Please let me know if I am missing something.
> > 
> > pranith @ /etc/glusterd
> > 17:12:09 :) $ sudo gluster volume create pranith `hostname`:/tmp/1
> > Creation of volume pranith has been successful. Please start the volume to
> > access data.
> > 
> You might want start volume here after creation and then stop it.  Then try
> add-brick again, even if that didn't start nfs server you can close the bug as
> Invalid.

Its working fine even with those steps. Marking the bug as works for me. Please re-open the bug if you see it later.

pranith @ /etc/glusterd
11:13:06 :) $ sudo gluster
gluster> volume create pranith pranith-laptop:/tmp/1
Creation of volume pranith has been successful. Please start the volume to access data.

pranith @ /etc/glusterd
11:13:34 :( $ sudo gluster volume start pranith
Starting volume pranith has been successful

pranith @ /etc/glusterd
11:13:48 :) $ ps aux | grep gluster
root      5658  0.6  0.4  50024  9108 pts/0    Sl+  11:13   0:00 glusterd --debug
root      5686  0.1  0.6  79836 13116 ?        Ssl  11:13   0:00 /usr/local/sbin/glusterfsd --xlator-option pranith-server.listen-port=24009 -s localhost --volfile-id pranith.pranith-laptop.tmp-1 -p /etc/glusterd/vols/pranith/run/pranith-laptop-tmp-1.pid --brick-name /tmp/1 --brick-port 24009 -l /usr/local/var/log/glusterfs/bricks/tmp-1.log
root      5692  0.5  1.7  93940 34900 ?        Ssl  11:13   0:00 /usr/local/sbin/glusterfs -f /etc/glusterd/nfs/nfs-server.vol -p /etc/glusterd/nfs/run/nfs.pid -l /usr/local/var/log/glusterfs/nfs.log
pranith   5701  0.0  0.0   7624   904 pts/1    S+   11:13   0:00 grep --color=auto gluster

pranith @ /etc/glusterd
11:13:50 :) $ sudo gluster volume stop pranith
Stopping volume will make its data inaccessible. Do you want to Continue? (y/n) y
Stopping volume pranith has been successful

pranith @ /etc/glusterd
11:14:00 :) $ ps aux | grep gluster
root      5658  0.5  0.4  50024  9108 pts/0    Sl+  11:13   0:00 glusterd --debug
pranith   5708  0.0  0.0   7624   904 pts/1    S+   11:14   0:00 grep --color=auto gluster

pranith @ /etc/glusterd
11:14:01 :) $ sudo gluster volume add-brick pranith pranith-laptop:/tmp/2
Add Brick successful

pranith @ /etc/glusterd
11:14:15 :) $ ps aux | grep gluster
root      5658  0.4  0.4  50024  9124 pts/0    Sl+  11:13   0:00 glusterd --debug
pranith   5715  0.0  0.0   7624   908 pts/1    S+   11:14   0:00 grep --color=auto gluster

pranith @ /etc/glusterd
11:14:17 :) $


Note You need to log in before you can comment on or make changes to this bug.