Re: [Gluster-users] Evergrowing distributed volume question

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

 



As gluster does not have metadata server, the client:s identify the brick via a special algorithm based on the file/dir name.

Each brick corresponds to a 'range' of hashes , thus when you add a new brick, you always need to rebalance the volume.

Best Regards,
Strahil Nikolov

Hello,

A while ago I attempted and failed to maintain an "evergrowing" storage
solution based on GlusterFS.
I was relying on a distributed non-replicated volume to host backups and
so on, in the idea that when it was close to full I would just add
another brick (server) and keep it going like that.
In reality what happened was that many of the writes were distributed to
the brick that was (in time) full, ending up with "out of space" errors,
despite having one or more bricks with plenty of space.

Can anyone advise whether current Glusterfs behaviour has improved in
this regard, ie does it check if a brick is full and redirect the
"write" to one that is not?

Regards,
Lucian
________



Community Meeting Calendar:

Schedule -
Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC
Bridge: https://meet.google.com/cpu-eiue-hvk
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-users
-------

Community Meeting Calendar:
Schedule -
Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC
Bridge: https://meet.google.com/cpu-eiue-hvk

Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-devel


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

  Powered by Linux