Bug 1162060

(In reply to Sachidananda Urs from comment #0)
> Description of problem:
> 
> When gstatus is run without any options, the capacity displayed is wrong.
> The (raw used) field constantly displays 12G.
> 
> [root@guido ~]# gstatus
>  
>      Product: RHSS v4            Capacity: 293.00 GiB(raw bricks)
>       Status: HEALTHY                       12.00 GiB(raw used)
>    Glusterfs: 3.4.0.68rhs                  146.00 GiB(usable from volumes)
>   OverCommit: No                
> 
> [root@guido ~]# df -h /rhs/brick1/
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/mapper/RHS_vg1-RHS_lv1
>                        81G  3.0G   78G   4% /rhs/brick1
> [root@guido ~]# 
> 
> On client:
> 
> [root@bob-the-minion ~]# df -h /mnt/gluster
> Filesystem            Size  Used Avail Use% Mounted on
> 10.70.34.100:gs       131G  6.0G  125G   5% /mnt/gluster
> 
> 
> Version-Release number of selected component (if applicable):
> 
> [root@guido ~]# gstatus --version
> gstatus 0.62
> 
> 
> How reproducible:
> 
> Always.

I'll take a look. I'm also interested in your product identification showing RHSS v4?

Created attachment 956392 [details]
gstatus output

sac, can you provide a gstatus -al from your environment please?

On my test environment the numbers are correct - gstatus -al, shows how the bricks are consumed, and a gstatus shows the same numbers aggregated. I've attached the output of the two commands for reference.

Paul, I think looking at -al makes more sense. It is the aggregation of the disk used on the bricks.

[root@guido ~]# gstatus -al
 
     Product: RHSS v5            Capacity: 293.00 GiB(raw bricks)
      Status: HEALTHY                       24.00 GiB(raw used)
   Glusterfs: 3.4.0.70rhs                  146.00 GiB(usable from volumes)
  OverCommit: No                

   Nodes    :  4/ 4             Volumes:  1 Up
   Self Heal:  4/ 4                       0 Up(Degraded)
   Bricks   :  4/ 4                       0 Up(Partial)
   Clients  :     1                       0 Down

Volume Information
        gs               UP - 4/4 bricks up - Distributed-Replicate
                         Capacity: (8% used) 12.00 GiB/146.00 GiB (used/total)
                         Self Heal:  4/ 4
                         Tasks Active: None
                         Protocols: glusterfs:on  NFS:on  SMB:on
                         Gluster Clients : 1

        gs-------------- +
                         |
                Distribute (dht)
                         |
                         +-- Repl Set 0 (afr)
                         |     |
                         |     +--10.70.34.100:/rhs/brick1/gs0r0(UP) 6.00 GiB/81.00 GiB 
                         |     |
                         |     +--10.70.34.101:/rhs/brick1/gs0r0(UP) 6.00 GiB/81.00 GiB 
                         |
                         +-- Repl Set 1 (afr)
                               |
                               +--10.70.34.102:/rhs/brick1/gs0r1(UP) 6.00 GiB/50.00 GiB 
                               |
                               +--10.70.34.103:/rhs/brick1/gs0r1(UP) 6.00 GiB/81.00 GiB 



Status Messages
  - Cluster is HEALTHY, all checks successful

However the Product field, RHS V5, I would rather like to see something like.

RHSS 2.1 V5.

The output of below command.

[root@guido ~]# cat /etc/redhat-storage-release 
Red Hat Storage Server 2.1 Update 5

+1

I've updated the getVersion function so now you get

[root@rhs3-2 ~]# gstatus 
 
     Product: RHSS v2.1 u4       Capacity:  80.00 GiB(raw bricks)
      Status: HEALTHY                        4.00 GiB(raw used)
   Glusterfs: 3.4.0.68rhs                  200.00 GiB(usable from volumes)
  OverCommit: Yes               

There's a 0.62-1 rpm in the repo now.

Of course reporting on capacity available is now (v3.x) problematic since thin provisioning is supported, and gluster itself doesn't see this layer. All the data in the tool comes from gluster so, if it doesn't see it, neither will gstatus.

Let me know if you need anything else - if not feel free to close the BZ.

Paul,

Thanks. Tested on 

# gstatus --version
gstatus 0.62

Version is fixed. Marking the bug as verified, I'll be creating a tracker bug to track all the Capacity related issues.