mastodon.ie is one of the many independent Mastodon servers you can use to participate in the fediverse.
Irish Mastodon - run from Ireland, we welcome all who respect the community rules and members.

Administered by:

Server stats:

1.6K
active users

#homelab

78 posts68 participants5 posts today

Credentials shouldn't be around in plain text files. But I also don't want to set up a fully fledged credentials management solution for my homelab.

Wouldn't it be nice to dynamically load the credentials I need when I step into my work directory, and remove then when I leave it?

Let's use @bitwarden and direnv to keep credentials safe in all simplicity!

ergaster.org/posts/2025/07/28-

ergaster.org · Loading credentials from Bitwarden with direnv
More from Thib

Added a few new gists on setting up a homelab Certificate Authority (CA) on a RHEL 10 machine with step-ca as podman container in preparation for a longer blogpost on the topic.

- Basic Step CA setup as podman container
- Manually add a root CA certificate to RHEL 10
- Manually generate certificates with Step CA

codeberg.org/jwildeboer/gists/

Tomorrow I will add a gist on using certbot to renew certificates in my homelab using that CA.

Summary card of repository jwildeboer/gists
Codeberg.orggistsA collection of short notes on specific little things that are good to store and share but not enough for a blog entry. Mostly geeky stuff.

I should start organizing the backlog for my home lab. First up will be figuring out what hypervisor I want to run; then some permanent infrastructure that won't be containerized such as a git server and pipeline runner; then making a VM template; then actually standing up the prod cluster.

Then comes the "migration" from my current instance to the new k8s cluster. That will be some combination of k8s deploys from YAML files and restoring some files from backup.

I plan to have a way to deploy new VMs on demand, including randomized hostnames and DNS entries, then assemble a whole group of them for k8s. Maybe I can even tinker with Ansible's dynamic groups?! Haven't done that for a while.

I'm looking for self-hosted alternatives for my Google Photos backups. I have some old photos on an external drive and I'd like to make them searchable, ideally within an android app. But I'd like to keep them within my local network. I could either simply share a folder via samba, e.g. Solid Explorer handles the network folders quite well, but I'm sure there must be better solutions. Recently, I found #immich, which looks very promising. #photoprism seems interesting. Any other ideas?

Replied in thread

@daj @joel @gotosocial Thanks for creating TagPush, David! I appreciate the work you're putting into solving the small instance discovery problem.

You're absolutely right about FediBuzz being erratic - I've noticed that too. Some days are content-heavy, others pretty quiet.

I actually tried both services and found different strengths:

TagPush: More reliable/consistent, but limited tag selection (totally understandable for a one-person project!)

FediBuzz: Broader tag coverage (I need things like #proxmox, #docker, #fediadmin and also #nrw and other hashtags for local content), but yeah, the reliability can be hit-or-miss.

For my use case I ended up sticking with FediBuzz mainly because I need some specific tech/local tags that aren't available on TagPush yet. But I really like TagPush's consistent delivery approach - that "fire-hose" problem you mention with relays is real!

Question: Are you planning to expand the tag list, or keeping it intentionally curated to maintain quality? I'd love to see tags like #docker, #proxmox, #homelab if that fits your vision!

Really appreciate both services existing - competition/alternatives are always good for the ecosystem! 🚀

Another one from my self hosting adventures - ended up setting up my own IRC server a while back, with UnrealIRCd6 and Anope, plus a bunch of custom python bots handling trivia (AI based), weather, search...

Took some work to get everything running smoothly (ports, proxies, SSL headaches), but it’s all humming now 😊

If you're into classic chat, testing cool bots, or just want to drop by and hang out, let me know — happy to share the server.

Yesterday I noticed that the pool pump wasn't running. I checked the power consumption logs and found it hasn't been running for a few days. This is very bad! High risk of soup pool.

After fixing the issue with the pump, I set up an automation in #homeassistant : At 09:30 every day, check the power consumption of the pump. Send an alert to my phone if it's less than 1.5kW.

All thanks to #foss!

Very strange behavior from #TrueNAS Scale tonight. First time this particular thing has ever happened. It has completely stopped responding on it's main IP address. I have 4 different VMs all running on it (each has its own IP address). All the VMs are fine. They're running normally, they're reachable, they're running fine. But I can't get to the main IP for the main system via SSH, HTTPS, nothing.

So the physical box and an awful lot of the networking is running just fine. DHCP server says it renewed its IP address less than 24 hours ago. So I will have to go drag a monitor and keyboard into the garage to reboot it. Weird.

I'm just wildly pleased with the switch to #NixOS.

My Pi was super dead after the latest storm related power issues, even the external drive doesn't power up properly anymore.

So I hacked together a deploy script and repurposed the mini PC I wasn't actually using. A couple days later and now I've got #HomeAssistant back up in a container, centralized my admin and backed it up to version control.

"Maybe someday" automation items were accessible immediately. I love it.

Continued thread

And I am done :-)

Number of #bitnami container images in the #homelab cluster: 4

Number of #bitnami container images successfully removed: 3

Number of #bitnami container images to remove: 0

Last up was moving Harbor to the official Helm chart. I can't really remember why I didn't stick with this previously... IIRC I had some challenges with the install but anyway the official helm chart looks to have worked as a direct drop-in with the same DB and S3 Bucket

goharbor.io/docs/2.13.0/instal

Harbor – Deploying Harbor with High Availability via Helm project logo
HarborHarbor – Deploying Harbor with High Availability via Helm