Cluster two node, without DRDB, only with external storage

[doknet]

Hello, I wonder.
Promox 2.1
1) Do I need to explain the changes in the wiki http://pve.proxmox.com/wiki/Two-Node_High_Availability_Cluster When the storage is an external server via NFS?




2) if not necessary should we leave it as is the file?

root@www12:~# cat /etc/pve/cluster.conf
<?xml version="1.0"?>
<cluster name="CLUSTIUOG" config_version="2">
  <cman keyfile="/var/lib/pve-cluster/corosync.authkey">
  </cman>

3) Is it necessary, repeat even if the storage is in an external server changes in cluster.conf?

root@www12:~# cat /etc/pve/cluster.conf
<?xml version="1.0"?>
<cluster name="CLUSTIUOG" config_version="3">
 <cman keyfile="/var/lib/pve-cluster/corosync.authkey" two_node="1" expected_votes="1"></cman>
  </cman>

4) Is it necessary, repeat enve if the storage is in an external server, change in cluster.conf without DRDB.

  <clusternodes>
  <clusternode name="www11" votes="1" nodeid="1"/>
  <clusternode name="www12" votes="1" nodeid="2"/></clusternodes>
</cluster>

5) I understand that it is not necessary to configure <fencedevices>

6) Someone could spend a sample configuration of cluster.conf in case you have external storage only, eg via NFS. -
7) an example of cluster.conf with two storage, external storage, and other with device in DRDB.


From already thank you very much.

 

[inspirit1986]

Hello. I don't want to create the next similar topic then I ask my questions here.

Can I create 2 proxmox servers with shared storage (vm disks) on third server (over iscsi)?

If 1st server is going down, I want to start my virual machines on the 2nd server.

So, can I mgmt my servers from any proxmox nodes if one of it is going down? (AFAIK, this technologies known as "Multimaster")

Explain me, how can I introduce this in my infrastructure.

thx a lot.

 

[doknet]

What you you describe is the same architecture as I want.
Two HA server (so that when you drop one replace the other) with external storage. (The storage server will be a third or a Nass or the like via NFS or iSCSI)
So I ask if it is necessary to make any change that architecture in the cluster.conf file.
The wiki. http://pve.proxmox.com/wiki/Two-Node...bility_Cluster .. is referred to as the storage is DRDB local disk.
So my question is completely external when that changes or not in cluster.conf when only two server in HA. -

Hello. I don't want to create the next similar topic then I ask my questions here.

Can I create 2 proxmox servers with shared storage (vm disks) on third server (over iscsi)?

If 1st server is going down, I want to start my virual machines on the 2nd server.

So, can I mgmt my servers from any proxmox nodes if one of it is going down? (AFAIK, this technologies known as "Multimaster")

Explain me, how can I introduce this in my infrastructure.

thx a lot.

 

[doknet]

resolved
My question was what to do with the cluster.conf file when the external storage is via nfs with two nodes.
I made no change in the cluster.conf file. which was my doubt if one does not work with DRDB.
As the wiki http://pve.proxmox.com/wiki/Fencing
just ran the command
join fence_tool
To be permanently
nano / etc / default / redhat-cluster-pve
# ClusterName = ""
# Nodename = ""
# USE_CCS = "yes"
# CLUSTER_JOIN_TIMEOUT = 300
# CLUSTER_JOIN_OPTIONS = ""
# CLUSTER_SHUTDOWN_TIMEOUT = 60
# RGMGR_OPTIONS = ""
FENCE_JOIN = "yes"


Thank you very much, sorry for asking something explained by colaboraderes of Proxmox.