Cluster error while adding new node

[lDemoNl]

Hi guys! I faced a problem when adding a node to the cluster. The cluster consists of 7 nodes, and when adding a new node, a failure occurred, most likely corosync, and all cluster virtual machines were turned off.
PVE version of all nodes in cluster 7.1-10, PVE version of new node 7.2-4.
Information from syslog

 

[oguz]

hi,

* can you post the contents of your /etc/pve/corosync.conf file?

* are you using separate network link for corosync?

* do you have HA enabled?

 

[lDemoNl]

hi,

* can you post the contents of your /etc/pve/corosync.conf file?

* are you using separate network link for corosync?

* do you have HA enabled?

Thank you for your response!

1)

Spoiler: corosync.conf

logging {
  debug: off
  to_syslog: yes
}

nodelist {
  node {
    name: ala220pve01
    nodeid: 3
    quorum_votes: 1
    ring0_addr: 172.16.133.136
  }
  node {
    name: ala220pve02
    nodeid: 4
    quorum_votes: 1
    ring0_addr: 172.16.133.137
  }
  node {
    name: ala2530pve01
    nodeid: 1
    quorum_votes: 1
    ring0_addr: 172.16.133.134
  }
  node {
    name: ala2530pve02
    nodeid: 2
    quorum_votes: 1
    ring0_addr: 172.16.133.135
  }
  node {
    name: ala924bf01pve03
    nodeid: 5
    quorum_votes: 1
    ring0_addr: 172.16.133.138
  }
  node {
    name: ala924bf01pve04
    nodeid: 6
    quorum_votes: 1
    ring0_addr: 172.16.133.139
  }
  node {
    name: ala924bf02pve01
    nodeid: 7
    quorum_votes: 1
    ring0_addr: 172.16.133.153
  }
  node {
    name: ala924bf02pve09
    nodeid: 8
    quorum_votes: 1
    ring0_addr: 172.16.133.154
  }
}

quorum {
  provider: corosync_votequorum
}

totem {
  cluster_name: pve-int01
  config_version: 8
  interface {
    linknumber: 0
  }
  ip_version: ipv4-6
  link_mode: passive
  secauth: on
  version: 2
}

2) No
3) Yes

 

[oguz]

according to your corosync config the 8th node is already added into your cluster.

the reason for the VMs being turned off might be related to HA rules.

a possible cause for nodes having connection issues with corosync could be the lack of a separate link (since it could get congested)

also just to be sure i'd recommend upgrading and keeping all nodes at the same package levels for consistency

 

[lDemoNl]

according to your corosync config the 8th node is already added into your cluster.

the reason for the VMs being turned off might be related to HA rules.

a possible cause for nodes having connection issues with corosync could be the lack of a separate link (since it could get congested)

also just to be sure i'd recommend upgrading and keeping all nodes at the same package levels for consistency

What do you mean "separate link" for corosync? Because I use separate vLAN's (cluster and management)
And could you tell me more details about HA rules?

 

[oguz]

What do you mean "separate link" for corosync? Because I use separate vLAN's (cluster and management)

yeah i meant separate LANs. are they on different physical links? [0] (see network requirements)

And could you tell me more details about HA rules?

you could read our documentation chapter for it [1]

[0]: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_cluster_network
[1]: https://pve.proxmox.com/pve-docs/chapter-ha-manager.html

 

[lDemoNl]

yeah i meant separate LANs. are they on different physical links? [0] (see network requirements)


you could read our documentation chapter for it [1]

[0]: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_cluster_network
[1]: https://pve.proxmox.com/pve-docs/chapter-ha-manager.html

No, I use 2 link 10G in bond (LACP enabled)
this is my network configuration