Recent content by Joi Owen

I've resorted to using ansible to customize the /etc/cloud-init/templates/hosts.debian.tmpl instead, since both Amazon EC2 and Proxmox seem to go out of their way to make cloud-init even more intrustive than Canonical designed it to be. I'm beginning to really loath cloud-init.

I opened a support case for this exact question. Handling keys isn't supported with the HTTP media type. It only works if you actually copy the media to the remote server. I hacked up a set of scripts to tarball just enough of the media tree to get the offline keys working. One runs on the...

I checked one of my production hosts, and the two .h files that script is looking for are both present. There may be an issue with that script experiencing a compilation error with the test. I'd try enabling verbose or debug output on the configure command and rerunning it. Double-check the...

Review the contents of config.out; the error is from the 1_bd_first_part.conf script in ./configure.d, and it suggests your host is lacking some header files. Devs tend to forget not every production server has a complete set of kernel header packages installed.

I used a 6.3 proxmox test host on my desk to migrate some VMware VMs into Proxmox. I got them booting and running on my test host. I did a vzdump to export them, then sneakernet to get them into their target proxmox clusters. The VMs I imported into a Proxmox 6.4 cluster worked fine. The VMs...

We use DigiCert for our public-facing certs, but all of our proxmox are internal only, and are using certs internally signed by our windows CA (because reasons...) DigiCert has used a variety of intermediate signing certs for their commercial certificates, and I've had them signed (more than...

I think you're colliding with the different chipsets in the graphical heads on the two hosts. X11/wayland/whatever are tightly bound to the hardware detected when X11 was originally installed. Check the Display and Hardware settings on the Vm and verify they're presenting the same chipsets to...

Booting a cloned target is always a trick, since the MBR and grub setup tends to break between source and target when you're doing filesystem rsyncs. What I've done in the past is boot both the source and target from a rescue cd, enable their networking, then use dd | netcat to copy the source...

We see a fair bit of that behavior, one cluster (recently upgraded from 5.4/luminous to 6.4/nautilus) was nearly unusable. The gear was under warranty (Super Micros servers with SM-branded LSI 3108 controllers in JBOD mode) and it turned out to be some as-yet unidentified issue between the...

I've already cloned that puppy! I'm new to ansible, but the pointy haired bosses here love it so I'm building out some playbooks of the common stuff I have to do to administer our zoo of proxmoxen. Do you take pull requests?

I've learned over the years that a lot of bugs in security functions can be avoided by sanitizing your passwords to remove common shell metacharacters like $*!#[]()<> and &. It so very common for some dumb thing to accept a password string and then send it unquoted into another shell command...

I would be interested in seeing what you have, yes.

Talk about timing. I just encountered this installing on an Ubuntu 20.04 test vm and hit this issue. Your test package installed properly.

I'm being told to install and configure some Intel Optane in some Proxmoxen, and some how incorporate it as disk cache in a 4-node cluster. There is a rather large mssql server in there which I believe is the primary target management is wanting to improve. I've been doing some searching and...