[one-users] The joy of Gluster

Peter Harris doilooksensible at gmail.com
Thu Dec 4 04:21:12 PST 2014


Hi

I have just installed a test OpenNebula system to figure out how it all
hangs together, and see if I can use it for our
development/support/projects departments.

I currently have:
- on-ctrl
4Gb RAM
146Gb HD
Centos 6.5
OpenNebula 4.10.1, installed as per
http://docs.opennebula.org/4.10/design_and_installation/building_your_cloud/ignc.html
Gluster 3.5.2, with on-node1 gluster volume mounted on
/var/lib/one/datastores/0 (with datasores/1 linked to datastores/1)

-on-node1
16Gb RAM
500Gb HD
3TB HD, shared via gluster volume gv-storage, replicated on on-node2
Centos 6.5
OpenNebula 4.10.1, installed as per
http://docs.opennebula.org/4.10/design_and_installation/building_your_cloud/ignc.html
Gluster 3.5.2 configured as per
http://docs.opennebula.org/4.10/administration/storage/gluster_ds.html,
with on-node1 gluster volume mounted on /var/lib/one/datastores/0 (with
datasores/1 linked to datastores/1)

-on-node2
16Gb RAM
500Gb HD
3TB HD, shared via gluster volume gv-storage, replicated on on-node1
Centos 6.5
OpenNebula 4.10.1, installed as per
http://docs.opennebula.org/4.10/design_and_installation/building_your_cloud/ignc.html
Gluster 3.5.2 configured as per
http://docs.opennebula.org/4.10/administration/storage/gluster_ds.html,
with on-node2 gluster volume mounted on /var/lib/one/datastores/0 (with
datasores/1 linked to datastores/1)

Hopefully this is not a massively stupid configuration. I appreciated my
shared filesystems should be on some form of SAN/NAS, but everything I have
is begged, stolen or borrowed, and I have been given a smaller budget than
normal (usually 0).

My question is, if I follow
http://docs.opennebula.org/4.10/administration/storage/gluster_ds.html, at
the stage "Configuring GlusterFS Datastore", I create the ds.conf, set the
correct GLUSTER_HOST (using the same port listed on the web page) and
GLUSTER_VOLUME name.

When I run onedatastore create and then onedatastore list, the new
datastore, ID 100, shows as the same size as the datastore directory on the
ctrl server
 onedatastore list
  ID NAME                SIZE AVAIL CLUSTER      IMAGES TYPE DS       TM
   0 system                3T 95%   -                 0 sys  -        shared
   1 default               3T 95%   -                 0 img  fs       shared
   2 files                79G 95%   -                 0 fil  fs       ssh
 100 glusterds            79G 95%   -                 0 img  fs       shared

I am confused, I was expecting glusterds to be the same as the 0 System /1
default datastores. Do I actually need glusterds, or can I just go ahead
and create my images on 1 default and have them used by the hosts (which
have the gluster volume mounted on /var/lib/one/datastores/0 ).

I have searched through the mail archive, and tried to follow what has been
said there, but I still have a confused.

Thanks in advance
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20141204/e815838e/attachment-0001.htm>


More information about the Users mailing list