⚔️ su || cd ⚔️

suorcd@getalby.com

cloudMaker || Podverse SysAdmin/DevOps || DevOps consultant || linux nerd || !dudeNamedBen || ham || low voltage storage

Public Key

npub1yda57lnlawz4qt78z9jjl0wul8nxx6hvn3xt6cvqjtywvnduc6wqvch4tz

NIP-05 Address

suorcd_at_podcastindex.social@momostr.pink

Profile Code

nprofile1qqszx7600el7hp2s9lr3zef0hhw0nenrdtkfcn9avxqf9j8xfk7vd8qpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dse7lzlf

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1yda57lnlawz4qt78z9jjl0wul8nxx6hvn3xt6cvqjtywvnduc6wqvch4tz

Show more details

Last Notes

2026-10-23 16:18:48 UTC

- reply

yes, this has been a week full of bugs and oddities...though I have not had to deal with it like others.

2026-10-23 16:14:36 UTC

Found a bug in my HA that has been lurking for years, but for some reason only showed up in the past couple days.
It is now fixed.
#homelab #yaml

2026-10-07 02:14:26 UTC

- reply

who said more?

2026-08-31 18:22:28 UTC

@npub1rcn…dw4h new #gossipwatcher PR https://github.com/Podcastindex-org/podping-gossipwatcher/pull/8
the changes you pushed to main, have resulted in no crashes over 33 hours.
@npub1z6g…amwj this PR does implement the cross-compile in a different way and adds armv7
#podping

2026-08-29 15:59:03 UTC

- reply

I'm working on getting my patch set cleaned up to 'rebase' on the new main.
I need to do a better job at branch management. right now the change set and the oom work are on the same branch.

2026-08-29 15:57:43 UTC

- reply

i'm not sure how that would fix the issue. I figured it was better to have unique keys vs. 3 containers with the same key.
if each container had the same key derivation i would think they would all create the same key.

2026-08-28 23:44:40 UTC

- reply

according to the bot each note is crashing about four times an hour, and I have three nodes in K3s

2026-08-28 23:40:21 UTC

- reply

with the way I have it set up, they are a replica set. Due to that fact, secrets, or config didn’t make sense to me. It seemed much easier just to let them spin up a new secret every time.

2026-08-28 19:33:16 UTC

- reply

the K3s containers rotate their key. Every time they die , due to the fact they have no durable storage.

2026-08-20 15:17:09 UTC

- reply

I still need to verify, but the bot did some digging and found no issues. Here’s the ticket it made https://github.com/suorcd/podping-gossipwatcher/issues/7

2026-08-19 14:56:12 UTC

- reply

I'm happy to test what is changed. For me it would be easiest if the changes were against https://github.com/suorcd/podping-gossipwatcher/tree/feature/builds02 since my deploy pipeline is already setup to pull the different architectures.
If that is too much i can make changes to my pipeline
#podping #gossipwatcher

2026-08-19 14:37:36 UTC

- reply

here is the bot analysis https://github.com/suorcd/podping-gossipwatcher/issues/3#issuecomment-5343601561
If we are not caring about this running on nodes with 1gig of ram or less we can backlog this and I can remove the RPI with 1gig and adjust the LXC to have more than 1gig of ram.
#podping #gossipWatcher

2026-08-19 00:07:06 UTC

- reply

the LLM says it’s the new version of iroh causing the issues. Which leads to my question, what are the minimum specs this should run on?

2026-08-18 18:12:51 UTC

- reply

I have a new build based on this issue https://github.com/suorcd/podping-gossipwatcher/issues/3
i did the 'smart' thing and put a `cargo update` on top of that. we'll see how these new images run on strictly no more than 1gig of ram
#podping #gossipWatcher

2026-08-11 16:09:04 UTC

- reply

might be the lxc container is too small, but it seems to continue the with the same issues
AI crash findings
Aggregated: https://gist.github.com/suorcd/b4b72275bdad8fd714aeaa0cd61c04a7
individual
https://gist.github.com/suorcd/052df50747562cdbe584f7d4707fc4bc
https://gist.github.com/suorcd/cd8ab2beba01f9fa2790f27396c74b71
https://gist.github.com/suorcd/0e95bf0d4651321a296e54c88e11234a
https://cdn.masto.host/podcastindexsocial/media_attachments/files/117/077/601/835/481/312/original/66dc32791384c115.png

2026-08-10 20:16:32 UTC

- reply

yes building with that flag https://github.com/suorcd/podping-gossipwatcher/blob/ea8445095ff7b4bca6865e127bd97691b0c6f1db/Dockerfile#L14

2026-08-09 18:09:10 UTC

- reply

Seems to be a memory leak. I he these graphs from my lxc containers not the PIs or linode VPS. That said this in in the binary and the OCI containers have not crashed.
The binaries get built first and then injected into a fresh container with my build process.
#podping #gossipwatcher
https://cdn.masto.host/podcastindexsocial/media_attachments/files/117/066/869/626/375/465/original/414be530e4ce7248.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/117/066/871/186/459/693/original/3031704ef3bb9cdf.png

2026-08-08 05:24:54 UTC

- reply

I had a couple x86 crash too. I wiped 3 of the 4 crashing data dirs; to give them a clean place to write data. At this moment they all are running; for at least 9 hours.
#podping #gossipWatcher

2026-08-07 18:40:06 UTC

The armv7 binary has not been as stable as they other, so my PR is not ready just yet.
I need to do some debugging, but my current setup does not keep logs. I'm working on adjusting that.
#podping #gossipwatcher

2026-08-07 18:24:24 UTC

- reply

my forth coming PR should have both of those architectures covered in binary and container format

2026-08-07 15:30:54 UTC

- reply

in my current PR. I’m actually working on separate binaries that are multi arch and multi arch containers.
I have been running two different versions of arm binaries on my Pi; though not a raspberry pi 2

2026-08-02 19:20:00 UTC

@npub1rcn…dw4h i've integrated the changes into my branch here is the PR
https://github.com/Podcastindex-org/podping-gossipwatcher/pull/1
Going through final checks right now.
We can change the version in my code if it looks good to you.
#podping #gossipWatcher

2026-08-01 22:25:46 UTC

- reply

the branch https://github.com/suorcd/podping-gossipwatcher/tree/feature/builds00 is almost ready for a PR.
When were you thinking of dropping the iroh change?
#podping #gossipWatcher

2026-07-28 22:42:38 UTC

@npub1rcn…dw4h i'm working a a branch to bring over some of the other builds
https://github.com/suorcd/podping-gossipwatcher/tree/feature/builds00
it is not fully tested yet, but initial tests are good.
#podping

2026-07-22 00:32:07 UTC

- reply

careful you’ll soon want to run your oven mesh vpn like nebula or even straight wireguard

2026-07-17 14:31:19 UTC

- reply

yep

2026-07-17 14:16:31 UTC

- reply

I thought I was there to change active text input fields

2026-06-25 15:26:30 UTC

- reply

reviewing the docker logs. Note the host is on linode as well.

2026-06-17 19:47:00 UTC

- reply

If the AI training computer is running in Dave's house it needs A/C to be stable; and the funds should come from the index.
I went to college and got my CS degree, but I ended up better at infra. So as of late AI codes and I understand the parts i care about. I use the tool to make what I want, because I don't get caught in the parts I don't care about; or to be honest, don't have time to focus on.
Also, I do let AI touch the code, but usually have it create PRs using git worktrees. A

2026-06-11 03:18:41 UTC

- reply

With the idea that #PVNG will support self-hosted deployments via k8s/k3s, it seems like the idea that each deployment would need to sign requests seems onerous.
cc: @npub1p7v…qql0

2026-06-11 01:47:12 UTC

Built my own foot gun today. I previously set up a script to restart #Tailscale after system suspend to keep DNS working. The catch? Whenever I explicitly turned Tailscale off, the script would silently turn it back on after suspend.
I found this out while migrating to Nebula. The transition is only partly done, so it looks like I'll just turn off the script and disable Tailscale's DNS while I still need the vpn.
#homelab

2026-05-07 21:51:41 UTC

- reply

this deploy uses proxmox, k3s, cloudflare tunnels, garage s3 compatible storage, and tailscale. @npub1p7v…qql0 has been working super hard on getting the base layer working so we have a great base layer.
#pvng

2026-05-07 20:38:48 UTC

- reply

one more step to a personalized podcast app experience.

2026-05-07 20:37:55 UTC

I still need to brand this and there is a need for a lot of polish...but here a personal deploy of #PVNG
https://pv.suorcd.us/podcasts?type=global&sort=recent
I just grabbed the top 1000 from the #podcastindex so not taking responsibility for what is up.
#podverse #homelab #k3s

2026-04-29 00:41:52 UTC

- reply

if we’re gonna respond with what you didn’t ask for. I daily drive the Ferris/sweep. Though I did source the parts, ordered the PCBs, solder and program it myself.
https://keebmaker.com/products/ferris-sweep

2026-04-26 16:05:00 UTC

- reply

yes. if i were to have the same pub key on 2 nodes what would the system do?

2026-04-25 17:15:24 UTC

@npub1rcn…dw4h what would happen if a #gossiplistener node shared a key with another node?
#podpingalpha

2026-04-22 22:16:30 UTC

- reply

#podverse has had great success with #zfs with on-the-fly-compression for our DB storage backend. We're sitting at a compression ratio of 1.87x

2026-04-17 00:04:17 UTC

@npub1rcn…dw4h what is Nbr ?
4294967295
seems high
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/417/115/914/215/283/original/a37be4a3df00fbed.png

2026-04-09 16:30:11 UTC

- reply

I will probably request to have that in the future. This sort of thing was part of my day job before I was laid off, so I have experience with these applications and how to manage them....or I did.
cc: @npub1rcn…dw4h @npub1tk4…3wu7

2026-04-09 16:14:29 UTC

- reply

hmmmm...the only one i can see that is a member of the Podcastindex-org Github Org is
Alberto Betella
albertobeta
@npub1tk4…3wu7 is this you?
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/375/627/658/897/993/original/213a96e8ba92e5d5.png

2026-04-09 15:43:28 UTC

- reply

currently it a manual process to sync my branch with upstream and that kicks off the build; and if i create a tag starting with 'v' and push it will also create a new GH action run.
If i get admin access in the Github PI org I can work on those builds.
#podpingalpha

2026-04-09 15:31:54 UTC

- reply

that has never worked, at least for me, and I do not have access to fix it.
docker pull ghcr.io/suorcd/gossip-writer:latest
docker pull ghcr.io/suorcd/gossip-listener:latest

2026-04-08 22:31:02 UTC

@npub1rcn…dw4h how long until the old nodes are no longer present?
My k3s nodes generate a fresh key every time there is a new deploy.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/371/449/705/665/553/original/fd9570cd79e61113.png

2026-04-07 22:58:06 UTC

- reply

deploying.

2026-04-07 20:22:40 UTC

- reply

nope. the command i run
tmux new-session -d -s gossip-listener "bash -c \"NODE_FRIENDLY_NAME='su || cd >> pp01' ARCHIVE_ENABLED=1 CATCHUP_ENABLED=1 ~/bin/gossip-listener\""
starts a tmux session just for the #gossiplistener

2026-04-07 20:09:23 UTC

- reply

The LLM says we should add tikv-jemallocator and jemalloc-ctl crates to Cargo.toml for better leak information. I'm assuming you already have something, but if you don't I can add the request to my issue.
#gossiplistener #podpinalpha

2026-04-07 19:59:35 UTC

- reply

https://github.com/Podcastindex-org/podping.alpha/issues/61

2026-04-07 16:23:54 UTC

- reply

trust, but verify ;)

2026-04-07 16:22:34 UTC

- reply

yes it has been happening my ram constrained LXC nodes as well
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/364/341/979/397/826/original/e2a8e3599aa9934b.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/364/341/986/301/027/original/05757fcf9b6e4453.png

2026-04-07 05:16:37 UTC

- reply

that’s what the binary builds are for…though I can’t test them all.
https://github.com/Podcastindex-org/podping.alpha/releases

2026-04-06 20:50:13 UTC

- reply

did you compile it or use the binary on Github?
cc: @npub1rcn…dw4h

2026-04-05 14:44:56 UTC

- reply

ok, that makes things easier.

2026-04-05 14:44:31 UTC

- reply

I want the images to be available on both platforms, and I will request the token soon, hopefully.
I still haven’t seen if the container images build on main repo. I’m guessing that they are private and just need to be made public.

2026-04-05 14:40:38 UTC

- reply

I will look into it, should be trivial now that we have the first container.
I’m not sure if it will be another binary in the same container image or another container image.

2026-04-04 20:15:09 UTC

@npub1rcn…dw4h could #pimonitor use the new
#gossiplistener as a feed source, instead of the index?

2026-04-03 19:09:48 UTC

- reply

yes.

2026-04-03 19:02:53 UTC

@npub1rcn…dw4h
seeing this issue
https://gist.github.com/suorcd/93d9fd7ce3ffea6b0b7441e42abd3dc4
#gossiplistener #podpingalpha

2026-04-01 17:09:06 UTC

- reply

what is the expected time for this? and is there a timeout counter that'll force the program to close?
most things i see on shutdown allow no longer than 90 seconds.
#gossiplistener #podpingalpha

2026-04-01 17:05:26 UTC

- reply

no i wanted to swap the binary for another build.
So went to thru the normal process and came across that behavior.
it is unexpected to not have something be completed shutting down in less than 1 minute.
Also in the issue it show that the process was reporting `[HEALTH] OK ` which is also not expected.
#gossiplistener #podpingalpha

2026-04-01 16:56:56 UTC

- reply

did it again
This in on my nixOS system and it built the binary
https://github.com/Podcastindex-org/podping.alpha/issues/48
#gossiplistener #podpingalpha

2026-04-01 16:49:24 UTC

- reply

i've deployed 0.4.2
had an issue where it was not quitting
killed it after 2 minutes with pkill
#podpingalpha #gossiplistener

2026-03-30 21:24:51 UTC

- reply

at the moment, i will keep it in mind. not planning on using it unless directed.

2026-03-30 20:32:54 UTC

- reply

just got this deployed. have not gotten far enough to check the output. That said it sure spits out a lot messages on killing with ctrl-c
#podpingalpha #gossiplistener

2026-03-30 20:31:45 UTC

- reply

i did not have that on.
didn't know it was an option
#gossiplistener #podpingalpha

2026-03-30 16:54:14 UTC

- reply

honestly it is for #gossiplistener I figure with the rapid development, it would be easier/better to update a single file and have the scripts deploy.
so not strictly supply chain concerns, but will help with that
in the past i would use JenkinsCI for this, but I do not have one running locally.
That said, it would have been a much quicker process to use Jenkins for this...not sure it would be better though.

2026-03-30 16:25:25 UTC

With my local gitea and gitea-act_runner testing out how to deploy binaries via commits.
#homelab #gitops

2026-03-29 23:46:43 UTC

- reply

surprisingly no
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/315/125/244/130/621/original/465d386671175d5d.png

2026-03-29 16:22:48 UTC

@npub1rcn…dw4h when an old version fails do you still want info about it?
Like this memory usage from v0.3.4
#gossiplistener #pp2alpha #podpingalpha
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/313/380/105/067/833/original/d313af25e262da27.png

2026-03-28 18:38:59 UTC

- reply

I'm running the various builds on my machines
I made a list of where my gossip-listeners
https://z.suorcd.us/services/gossip-listener/
#gossiplistener

2026-03-28 18:37:01 UTC

@npub1rcn…dw4h PR ready https://github.com/Podcastindex-org/podping.alpha/pull/34
#gossiplistener

2026-03-28 18:01:43 UTC

- reply

speed of AI i just upgraded to 0.3.3 like 20 minutes ago and already out of date.

2026-03-28 17:01:33 UTC

- reply

working on PR

2026-03-28 16:24:16 UTC

- reply

do you mind if the Build x86_64-pc-windows-msvc is removed.
It was added by #vibe_coding and I can not maintain it, nor do i want to.
#gossiplistener

2026-03-28 02:37:32 UTC

- reply

2026-03-21 18:05:28 UTC

- reply

so far, with vibe coding, it seems easier than hugo. currently writing a custom theme
#zola

2026-03-21 17:33:57 UTC

moving my blog from hugo to #zola (written in rust)
mainly i'm hoping it will be more simple
I wanted to use #silverbullet but that loads a bit slowly for this.
https://gohugo.io/
https://www.getzola.org/
https://silverbullet.md/

2026-03-20 15:11:00 UTC

- reply

I also have
Binary being built as well for Linux, arm and Mac OSX though the ladder two haven’t been tested

2026-03-20 15:08:53 UTC

- reply

would it be possible to have a dashboard of active listeners. And it may be even a flag that says make their listening public?

2026-03-19 23:20:18 UTC

- reply

everything, 7 nodes, is running 0.2.5 now

2026-03-19 23:06:13 UTC

- reply

i've been busy with neighborhood IT. I'm building the images now

2026-03-19 16:03:40 UTC

- reply

yea and the Dell tool says there is no hardware issue

2026-03-19 15:17:18 UTC

I’ve been out for over six month.
My elderly neighbor has been having Windows 11 issues with his 3 computer; current one 6 months old. I have agreed to attempt to help.
His boot device is ‘inaccessible’ and he only has a passkey on his iPhone with access. My plan is to get the recovery key, backup the data and see if the Dell “Repair” button works.
As a person who gave up on Windows in 2010, is there any advicy from the crowd; I have been asking an LLM for specifics.

2026-03-18 16:00:51 UTC

@npub1rcn…dw4h
My post with the dtt error might have gotten traction
https://github.com/Podcastindex-org/podping.alpha/issues/20
#gossiplistener

2026-03-17 17:38:57 UTC

- reply

here is a PR https://github.com/Podcastindex-org/podping.alpha/pull/19
- add the Cargo.lock
- updates the docker container image build
- updates the flake.nix

2026-03-16 22:23:45 UTC

- reply

why was the Cargo.lock removed from #gossiplistener

2026-03-14 00:32:16 UTC

@npub1rcn…dw4h why is the gossip_listener_node.key data and not a string?
#p2alpha #gossiplistener

2026-03-13 22:37:57 UTC

- reply

is there a way to find out without restarting?

2026-03-12 22:28:34 UTC

- reply

and as soon as i post that is see that my k3s pods are getting podpings

2026-03-12 22:27:18 UTC

- reply

let me know when you want me to update/test.
#gossiplistener #podpingalpha #pp2alpha

2026-03-12 20:33:44 UTC

- reply

let me know if I can add another writer or more processing

2026-03-12 20:12:57 UTC

Is anyone else see
[WATCHDOG] No gossip notifications for 2880s, re-bootstrapping...
on #gossiplistener v0.1.8?
cc: @npub1rcn…dw4h

2026-03-12 19:32:57 UTC

- reply

this
```
[ENDORSE] PeerEndorse from untrusted sender e3bf2ce5, ignoring
```
still happening regularly. are you seeing it
#podpingalpha #gossiplistener

2026-03-11 17:45:03 UTC

- reply

you might want to add the --release flag. it might be a bit slower to build, but run better
`cargo run --release`
#gossip-listener #podpingalpha

2026-03-11 17:42:58 UTC

is anyone else seeing
```
[ENDORSE] PeerEndorse from untrusted sender e3bf2ce5, ignoring
```
in #gossip-listener
#podpingalpha

2026-03-11 17:31:00 UTC

@npub1rcn…dw4h i think the changes are solid enough to be helpful https://github.com/Podcastindex-org/podping.alpha/pull/9

2026-03-11 15:29:54 UTC

@npub1rcn…dw4h when you can, please take a look at this PR https://github.com/Podcastindex-org/pimonitor/pull/13

2026-03-08 02:39:26 UTC

- reply

yes there should be a place to grab some up to date gossip listeners.

2026-03-08 02:38:40 UTC

- reply

I have the k3s pods up and running too. they thus far do not have the startup issue.
And I'll try to get one in Linode Texas as well.

2026-03-08 02:31:04 UTC

- reply

i grabbed the data from one of the working gossip_listener_known_peers.txt and put it into the non-working nodes and that seemed to work