Resolved by setting network configuration to DHCP instead of static. Not sure what is wrong with the static configuration seeing as it's very minimal, but alas.
I also note that this only started happening after OVH had a network failure on a rack, affecting many servers. 3 of my servers were affected, 2 are fine now but 1 has this hourly network issue. They also replaced my motherboard and I think this may be the culprit but not sure how to diagnose it.
For some reason my host machine is losing internet connection every 30-60 minutes or such. When I run 'service networking restart' Proxmox is accessible again, but the VMs are not accessible on the internet until I shutdown and start the VM (a reboot does not solve).
'service networking status'...
No worries. In that case, if I have a compressed 'qcow2' and decide to full clone it to a 'raw', this should decompress all of the existing data and ensure any potential performance issues caused by compression to be void?
Many thanks!
Thanks for your clarification, however
According to one of the sources you provided https://www.jamescoyle.net/how-to/1810-qcow2-disk-images-and-performance it mentions that any compressed data will need to be decompressed as soon as it is accessed, so I assume this would degrade performance...
I have a qcow2 disk image that I compressed with "qemu-img convert -c". But I understand that this can inflict performance issues when running a compressed qcow2 disk image.
Thankfully I only plan to use the compressed qcow2 disk images as templates. So when I clone the compressed qcow2...
Solved by using qcow2 instead: https://pve.proxmox.com/wiki/Shrink_Qcow2_Disk_Files
All my Windows/Linux setups have trim setup automatically since installation, then enabling discard and converting to qcow2 without compression has saved space drastically. With compression saved even more space.
Merely because VMDK offers better compression, which is better suited when using inactive VMs as templates. I assume from your quote however that there is no better compression method, however please correct me if I'm wrong.
I am using raw for my VMs, but for templates I am trying to figure out what is the best way to get the most compressed filesize so that I can save more space for active VMs rather than have space used by templates. I am using '.vmdk' for templates currently but sometimes these still take full...
I followed this: https://pve.proxmox.com/wiki/Upgrade_from_5.x_to_6.0#In-place_upgrade
Then rebooted machine and tried to start VMs and this error shows:
/usr/bin/kvm: error while loading shared libraries: libtirpc.so.3: cannot open shared object file: No such file or directory
command...
Sorry to bump up an old thread, but I've not been able to accomplish this. What is the 'storage view' permission you gave to solve this?
I performed the following command:
pveum roleadd testuser --privs "VM.Audit VM.Backup. VM.Config.CDROM VM.Console VM.PowerMgmt Datastore.Audit"
But the user...
I can't afford €959,60 annually just to fix a small issue such as this, I was hoping I could find someone that would help me fix it for a one-time price.
Perhaps I could PM the link to you if you would like to test it?
If it helps, sometimes the site will go to 'ERR_CONNECTION_RESET' on Chrome even though it still hangs, but just says 'This page cannot be displayed' on Internet Explorer.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.