Re: 'T' sticky bit set on some files after rebalance

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Also I should say that all bricks still have plenty of free space, but
there were network problems in the LAB..


On Tue 04 Mar 2014 09:26:23, Viktor Villafuerte wrote:
> I've found this
> 
> [root@xxxxxxxxxxxxx ~]# getfattr -dm . Alien3_1992_25_HLS_layer2_642000_282.ts 
> getfattr: Removing leading '/' from absolute path names
> # file:
> # Alien3_1992_25_HLS_layer2_642000_282.ts
> trusted.gfid="��\012�h�Bȹ!v�/�,�"
> trusted.glusterfs.dht.linkto="cdn-uat-replicate-0"
> 
> [root@xxxxxxxxxxxxx ~]#
> 
> From some smart posts I found that this happenes when a brick fills and
> the file is then placed on another brick. The extended attr then points
> to the file's new location.
> 
> This would point to the 'cdn-uat-replicate-0' pair of bricks, however
> the files are not there. They are on another pair or replicated bricks.
> 
> cdn-uat-replicate-0 is in fact gluster0[8,7].uat
> 
> The actual files are located on 'cdn-uat-replicate-3' which is
> gluster0[5,6].uat
> 
> 
> Now, the actual Gluster mount seems ok, the files do exist and there's
> nothing I can do to fix this with heal, rebalance and friends.
> 
> I think I'm just gonna delete them..
> 
> 
> v
> 
> 
> 
> 
> 
> On Sat 01 Mar 2014 17:38:01, Viktor Villafuerte wrote:
> > Hi all,
> > 
> > after rebalance with distributed replicated volume setup I'm left with
> > files with sticky bit set. I have currently 3 replicas and only two of
> > them contain these files. They only appeared after second rebalance 2 ->
> > 3 replicas. 1 -> 2 replicas rebalance finished with all files not having
> > the sticky bit set
> > 
> > 1 x (1x1) => 2 x (1x1)
> > rebalance completed ok
> > no files with --------T set
> > 
> > 2 x (1x1) => 3 x (1x1)
> > rebalance completed ok
> > replica 0 - 95 files with 'T'
> > replica 1 - 0 files with 'T'
> > replica 2 - 265 files with 'T'
> > 
> > eg
> > 
> > ---------T 2 1000 1000       0 Feb 28 23:53
> > ThePinkPanther2_2009_23_HLS_layer4_1450000_89.ts
> > 
> > 
> > 
> > somehow I don't think this is expected. Rebalance does not change
> > anything and all shows as things are ok... but are they?
> > 
> > Anybody has any pointers for this?
> > 
> > 
> > -- 
> > Regards
> > 
> > Viktor Villafuerte
> > Optus Internet Engineering
> > t: 02 808-25265
> > _______________________________________________
> > Gluster-users mailing list
> > Gluster-users@xxxxxxxxxxx
> > http://supercolony.gluster.org/mailman/listinfo/gluster-users
> 
> -- 
> Regards
> 
> Viktor Villafuerte
> Optus Internet Engineering
> t: 02 808-25265
> _______________________________________________
> Gluster-users mailing list
> Gluster-users@xxxxxxxxxxx
> http://supercolony.gluster.org/mailman/listinfo/gluster-users

-- 
Regards

Viktor Villafuerte
Optus Internet Engineering
t: 02 808-25265
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://supercolony.gluster.org/mailman/listinfo/gluster-users





[Index of Archives]     [Gluster Development]     [Linux Filesytems Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux