I do use some storage on the machine which provides the quorum for some simple backups. This is happening on clusters with and without storage located on the device providing a quorum disk. This was also dietmer's thoughts, I provided output showing that it didn't seem like it was hanging while the quorum disk machine was down. I am stumped at this point, but there is no doubt the cluster itself is 100% working as expected.
Quorum Disk down causes odd GUI behaviors
- Thread starter adamb
- Start date
I do use some storage on the machine which provides the quorum for some simple backups. This is happening on clusters with and without storage located on the device providing a quorum disk. This was also dietmer's thoughts, I provided output showing that it didn't seem like it was hanging while the quorum disk machine was down. I am stumped at this point, but there is no doubt the cluster itself is 100% working as expected.
I really am starting to question why we even bother to purchase subscriptions. The "Enterprise Repo's" have proven to be no more stable than the "Non Enterprise Repo's" and the dev's provide responses when they feel like it. This is clearly a proxmox issue which should be addressed. I do have the ability to create tickets, but imo I shouldn't be wasting our tickets on bug's within proxmox.
I have proven and stated time and time again that this is a proxmox issue and not a configuration/quorum disk issue.
Hi adam,
when this problem occur,
- what is the content of /var/log/cluster/corosync.log
- are you able to write into /etc/pve/ ? (for example, can you open a /etc/pve/qemu-server/vmid.conf and write it ?).
can you try to restart
/etc/init.d/pve-cluster restart
/etc/init.d/pvedaemon restart
/etc/init.d/pvestatd restart
when this problem occur,
- what is the content of /var/log/cluster/corosync.log
- are you able to write into /etc/pve/ ? (for example, can you open a /etc/pve/qemu-server/vmid.conf and write it ?).
can you try to restart
/etc/init.d/pve-cluster restart
/etc/init.d/pvedaemon restart
/etc/init.d/pvestatd restart
I doubt that you waste your time on the Support. So far you got a lot of info and hints to fix your "not optimal" network and cluster setup - if you are unhappy with the support services, write your complains to office@.. - the forum is the wrong place for this.
Yep you guys have assisted me in a few issues and I do appreciate it, but lets face it, they were nothing that serious. I never said I waste my time, I simply said I wonder why we purchase them when the dev's respond to issues when they feel like it. They asked for the output of a few things which I provided, then never responded back?
I don't see how my setup is not optimal, are we going to start playing that hole quorum disk is clunky game again when its a fully supported and certified setup per red hat engineers? I understand this is promxox and not red hat, but the majority of it is built off the same concepts.
I do agree this is not the place and I should have kept that remark to myself. At the same time, the majority of this thread is me trying to work through the issue and community members providing input.
I did take offence by the first remark from the proxmox dev because I feel instead of really trying to understand the issue he wanted to blame it on the quorum disk.
Wouldn't it make more sense to provide input and assistance on my issue instead of telling me were the complaints go?
Last edited:
I am the forum moderator and I have to make sure that all post are clean and to the forum topic. As you do off-topic comments, I commented on this. If you need direct contact to the developers, use the commercial support - here is a short explanation what you can expect from the different support channels:
http://pve.proxmox.com/wiki/Get_support
if you found a bug in the Proxmox VE code, just filed it on https://bugzilla.proxmox.com - no support ticktet needed.
Do not forget to add a testcase so that our devs can easily reproduce the bug in their test lab.
Do not forget to add a testcase so that our devs can easily reproduce the bug in their test lab.
Well, I went to test some of the suggestions and I am unable to reproduce the issue on any of my clusters now. Quite odd. I will update this thread again when it happens and I pin down exactly what is the cause. I appreciate everyone's input.