Re: Understand Geo Replication of a big Gluster

Axel Gruber <axel@xxxxxx> · Fri, 15 Jun 2018 21:29:01 +0200

Hello Thanks for Information
I tryed to install Geo Replication again - setup SSH Key - prepared session Broker and so on (shown in the Manual)

But i get this error:

root@glusters1:~# gluster volume geo-replication gpool geo.com::geovol create force
Please check gsync config file. Unable to get statefile's name
geo-replication command failed

I use "force" command because Slave Gluster is to small - but its empty - so whtiout Force i get:

root@glusters1:~# gluster volume geo-replication gpool geo.com::geovol create
Total disk size of master is greater than disk size of slave.
Total available size of master is greater than available size of slave
geo-replication command failed

Can anyone help me to understand whats going wrong here ?

Kontaktieren Sie mich direkt per Live Chat - einfach hier Klicken:
----
Mit freundlichen Grüßen
Autohaus A. Gruber OHG
Axel Gruber / Geschäftsführer

Tel: 0807193200
Fax: 0807193202
E-Mail: axel@xxxxxx
Internet: www.autohaus-gruber.net

Ihr starker MAZDA und HYUNDAI Partner 4mal in der Region - einmal auch in Ihrer Nähe.

  Autohaus A. Gruber OHG, Gewerbepark Kaserne
  10, 83278 Traunstein.
  HRA 8216 Amtsgericht Traunstein,
  Ust.-Id.-Nr. DE813812187
  Geschäftsführer: Axel Gruber, Anton Gruber
  Steuernummer: 141/151/51801

Am Fr., 15. Juni 2018 um 03:37 Uhr schrieb Kotresh Hiremath Ravishankar <khiremat@xxxxxxxxxx>:
Hi Axel,

No geo-replication can't be used without SSH. It's not configurable.
Geo-rep master nodes connect to slave and transfers data over ssh.

I assume you have created the geo-rep session before start.
In the command above, the syntax is incorrect. It should use "::" and not ":/"

gluster volume geo-replication gpool glusergeo::geovol start

Thanks,
Kotresh HR

On Fri, Jun 15, 2018 at 1:32 AM, Axel Gruber <axel@xxxxxx> wrote:
Hello
First of all Thank you for Information

As First Try i want to setup Geo -Replication in a Test System with Virtual Machines

So for the Test i have a distributed Gluster with 3 Nodes:

Status of volume: gpool
Gluster process                             TCP Port  RDMA Port  Online  Pid
------------------------------------------------------------------------------
Brick glusers1:/mnt/glust/gpool             49152     0          Y       568
Brick glusers2:/mnt/glust/gpool             49152     0          Y       574
Brick glusers3:/mnt/glust/gpool             49152     0          Y       665
NFS Server on localhost                     2049      0          Y       1127
NFS Server on glusers2                      2049      0          Y       926
NFS Server on glusers3                      2049      0          Y       695

on The Slave Side i have created also a  volume

Status of volume: geovol
Gluster process                             TCP Port  RDMA Port  Online  Pid
------------------------------------------------------------------------------
Brick glusergeo:/mnt/glust/geo              49152     0          Y       629

The Maser Side can "see" the Hostname and Ping the IP of the Slave
root@glusters1:~# ping glusergeo
PING glusergeo (172.20.230.86) 56(84) bytes of data.
64 bytes from glusergeo (172.20.230.86): icmp_seq=1 ttl=64 time=3.85 ms

So for Test i not want to use SSH and i Tryed:

root@glusters1:~# gluster volume geo-replication gpool glusergeo:/geovol start
Invalid Url: glusergeo:/geovol

Is it not possible to user geoReplication without SSH ?

Best Regards

Am Do., 14. Juni 2018 um 16:27 Uhr schrieb Milind Changire <mchangir@xxxxxxxxxx>:
Hello Axel,
A warm welcome to you and happy to see you board the Gluster ship.

geo-replication requires two clusters: one master cluster which is the source cluster and the second cluster called the slave cluster which is the destination cluster

Just as you have created a primary/master cluster, you need to create another cluster out of the backup servers in whatever configuration (volume type) you want to create. The volume type of the slave cluster need not be the same as the master cluster.

Once you've created the slave cluster, you can then use geo-replication commands to create a communication channel between the two clusters so that the sources from the master cluster are replicated to the slave cluster and to keep pushing any updates to the slave cluster.

Hope this helps.

On Thu, Jun 14, 2018 at 6:23 PM, Axel Gruber <axel@xxxxxx> wrote:
Hello 
Im new on GlusterFS - so a warm hello to everyone here...

Im testing GlusterFS since some Weeks in different configurations for a big Media Storage

Currently for start we plan a distributed / Replicated Gluster with for Nodes (4x70Tb)
I tryed this within a Test Area on differenzt Virtual Machines - works fine.

But for security Reason we also plan Geo Replication of the Whole Gluster. I have read Doc  - but one thing i dont understand:

For example our Start Gluster has 4 Nodes (4x70TB  Replicated and Distributed) i have 140TB Capacity.

So when i want to Geo Replicate this Volume i dont have a sigle Server wich is able to store 140GB (or later more) - but i can have several Backup Servers with a total Sice of 140TB

So my Question 

How can i say Gluster FS in Geo Replication to use all of this Backup Servers to GEO REplicate the whole Storage using all Backup Server.

Best Regards

_______________________________________________

Gluster-users mailing list

Gluster-users@xxxxxxxxxxx

http://lists.gluster.org/mailman/listinfo/gluster-users

-- 
Milind

_______________________________________________

Gluster-users mailing list

Gluster-users@xxxxxxxxxxx

http://lists.gluster.org/mailman/listinfo/gluster-users

-- 
Thanks and Regards,
Kotresh H R

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users