Echo Nar

@lethedata
2 Followers
18 Following
173 Posts
I just pretend to know what I'm doing.
- A tech nerd who loves cute adorable things!
- An IPv6 advocate who understands IPv4
Websitehttps://lethedata.com/
Codeberghttps://codeberg.org/lethedata
GitHubhttps://github.com/lethedata
At first I though OneUptime had some useful blog posts and guides. Then I realized they're just playing the SEO game by generating thousands of articles to appear in as many search results as possible. It's rather annoying.
I was having a good time making hierarchical IPv6 subnets... Then I had to deal with the IPv4 subnets 😢
Another day of LLM being a decent tool at one moment and completely useless at another.

Travis confirmed that the flags have higher preference so my cluster is running v6 :D

https://github.com/rook/rook/discussions/17214#discussioncomment-16215010

Is IPv6 bound even with ceph config ms_bind_ipv6 false? · rook rook · Discussion #17214

Is IPv6 the binding address even tho ceph config shows ms_bind_ipv6 as false? kubectl -n rook-ceph exec -it $(kubectl -n rook-ceph get pod -l "app=rook-ceph-tools" -o jsonpath='{.items[*].metadata....

GitHub

Sometimes I feel clever.

Wanted a way to pass k8s EncryptionConfiguration to hosts before the cluster was deployed but didn't like having to copy the file and then have two unencrypted copies. So what I did was wrote a script to generate the config file using TPM secured key from systemd-creds. This let me prepass the key securely to systems before the cluster was ready to be deployed.

mmm, I have no idea. I guess I'm just have to trust that with ROOK_POD_IP set to v6 and ms-bind-ipv6 flags it's actually using v6. I did ask on the github discussions page so we'll see.

Also it's completely possible all my issues getting mons started are stemming from "uncleaned" drives as sometimes it works and sometimes it doesn't. 🫠

or maybe not... Using flags over config I think
I'm like 90% sure rook-ceph dualStack and ipFamily IPv6 flags are bugged out. They don't enable ms_bind_ipv6 causing the mons to fail out leading to a single mon-a which ofc doesn't spin up the osds due to no quorum
Ok, got a tuned k0s to use IPv6 more than IPv4 in default dualstack. Now to bicker with ceph

With all this work to get k0s IPv6 dualstack primary running, I'm gunna cry when things on top just fail

(_゜_゜_)