Proxmox VE How To Setup High Availability

[@TechTutorialsDavidMcKone]

If you want to learn more about Proxmox VE, this series will help you out kzbin.info/www/bejne/qXm6ioiqZbtgmZo

[@Shimi1990]

The best proxmox HA Guide!!! thank you!

[@TechTutorialsDavidMcKone]

Thanks and good to know the video was helpful

[@mohamedalisahnoun8021]

briliant as usual !

[@TechTutorialsDavidMcKone]

Thanks for the feedback Good to know the video was useful

[@mohamedalisahnoun8021]

@TechTutorialsDavidMcKone though VMware was buyed by Broadcom licence are more expensive, do you recommend proxmox instead in production environment ?

[@TechTutorialsDavidMcKone]

@@mohamedalisahnoun8021 Well it is an Enterprise hypervisor, but as ever, it depends The hypervisor deployments I've seen over the years tended to be very basic i.e. just running some VMs on a cluster which do their own HA In those cases the hypervisor you use makes no difference But at most companies I've worked for, not many staff know their way around Linux For Proxmox, most things will be done in the GUI, but some have to be done in the CLI So being reasonably familiar with the Debian OS would help a lot, just in case To be fair though, even that's simple compared to when you installed Cisco's 1000V switches in ESXi

[@mohamedalisahnoun8021]

@TechTutorialsDavidMcKone one more question sire If I had one cluster with 3 nodes with a shared external storage iscsi or nfs ,if one node fails her vm configuration i can restore it to another node ?via etc/pve/nodes I just copy vm.conf Into another node I run it ??? Is this the best practice the disk is shared though

[@TechTutorialsDavidMcKone]

@@mohamedalisahnoun8021 If all of the nodes have access to the same storage then any of them can run the vm And servers in a cluster already synch the main folder so each has a copy of it

[@brainamess2979]

Really insightful thank you.

[@TechTutorialsDavidMcKone]

Good to know the video was helpful Thanks for the feedback

[@cooki3cutt3r13]

great info, thank you good sir

[@TechTutorialsDavidMcKone]

Good to know the video was helpful, so thanks for the feedback

[@kacperm2909]

very good material!

[@TechTutorialsDavidMcKone]

Thanks for the feedback

[@michaelcooper5490]

Another Great video David Awesome content thank you again for your insight.

[@TechTutorialsDavidMcKone]

Glad you liked it

[@jorgegomez374]

It's off topic to the video. I am building my promox server, I am using right now 1 pi and a Dell PC whit docker both in one network segment, a pi running home assistant whit diferente network segment my IoT network. The idea is to have everything on proxmox. My question is, do I need to add a nic to have the separation of my network? Thank you. I am following your videos for when I build it. Thanks

[@TechTutorialsDavidMcKone]

Yes you do need different interfaces to separate networks They don't need to be physical though You can create logical interfaces, if the operating system supports it, but you can miss out on bandwidth because it will be shared Check out my video for Proxmox and VLANs on a single NIC for instance kzbin.info/www/bejne/opvUZ6qir9Nnp9E

[@jorgegomez374]

@@TechTutorialsDavidMcKone thank you

[@maherkhalil007]

It looks great but I think proxmox does not support fail-over IP, so when you move VM still the IP allocated to the first node, I think VM will not be reachable?

[@TechTutorialsDavidMcKone]

The VM will keep it's IP address if it was assigned statically If it was assigned by DHCP, it will be retained most of the time It's usually desktop hypervisors that provide their own DHCP service to their VMs but they wouldn't be clustered like this to provide HA An enterprise hypervisor like PVE relies on a separate DHCP server to allocate IPs to VMs The risk to using DHCP is if the IP address expired and was allocated to another computer while the VM was offline. Highly unlikely but possible On some hypervisors the MAC address can change when a VM is migrated like this, which can then be a problem for reserved IPs in DHCP But with PVE I noticed the VM keeps the same MAC address So as long as the IP address is static or reserved there shouldn't be a problem

[@Darkk6969]

@@TechTutorialsDavidMcKone Yes that would be correct. Long as MAC address remains the same you're good to go.

[@vanhoatvdat]

it will automatically create new disk of VM on node 2? how it keep the disk data of VM on node 1 and migrate to node 2?

[@TechTutorialsDavidMcKone]

It depends A cluster only really works with shared storage In that case, the hard drive doesn't move when a VM is migrated as part of HA, although the contents of the RAM will need to be if the VM is live If a node fails, another node will reset the VM to take over the file lock(s) and there will be a loss of data if it was in RAM and wasn't written to disk I don't know the specifics of KVM for local storage transfers But one of the options I've noticed for migrating is to keep the original disk, so I assume the process is similar At a high level hypervisors copy the data from the existing hard drive to the new destination and then they start to keep the two drives in synch They do the same for the RAM if it's live Back when I was studying VMware for instance, it would keep the two VMs in sync and then there would be a brief pause and the old VM would be disconnected and the new VM would take over on the new node After that the older files would be deleted

[@mohemmedredha1424]

Sir i have 7 node cluster but network disconnect for node no6 and 5 and 2 whet it is back the HA show on these nodes old timestamp-dead? How to fix it plz?? and didi this effect on cluster

[@TechTutorialsDavidMcKone]

If you're not using HA for VMs then the timestamp message doesn't matter Otherwise you can check the log file on each server as that can provide more clues tail /var/log/syslog

[@mohemmedredha1424]

@ thank you very much But im using HA in my cluster also i have critical systems should not down How can I fix this issue plz Sir??

[@TechTutorialsDavidMcKone]

@@mohemmedredha1424 I suggest raising a ticket with Proxmox if it's critical

[@mohemmedredha1424]

@@TechTutorialsDavidMcKone im not have subscription so i need your help please should I restart these nodes thathave issue?? Or just restart cluster use commands to restart ?

[@TechTutorialsDavidMcKone]

​@@mohemmedredha1424 Unfortunately, that's not something I can help with I offer basic assistance if someone has difficulty repeating what was done in a video I've released For instance, maybe I missed a step or didn't explain things well enough in the video Maybe an upgrade has come out and what was shown in the video no longer works Anything above that is best served by the product vendor or another company offering IT support Because it may require a lot of time to investigate the problem and come up with a solution You can try and get assistance from community forums But the risk is someone suggesting doing something that makes a bad situation worse So if this is a production environment, it would be better to sign up to support from Proxmox