Full ALPR Database System for Blue Iris!

Hi all,

A few things have piled up over the past several months:

-I moved homes and don't have an active LPR setup to test against yet.

- I changed jobs

- I somehow developed a major endocrine disorder that causes a whole bunch of really gnarly issues, including significant cardiac stress. 167/107 BP when discovered, which is insane. Had no idea that this existed and is something that could just happen. Should be fine and not super relevant as a blocker here, but was a huge WTF.

- This is a good cautionary tale to not minimize or put off dealing with out-of-range results if u get bloodwork done. Had to learn the hard way on this one.


- I took on a project at a university on the east coast running from February through April, which involved a bunch of weekend travel flying back and forth.

- A separate commitment takes up most of my evenings through the end of June.

Ive almost tempted myself to go finish the code a few times as a mode of procrastination avoiding other actual obligations, but this is a single-maintainer open source tool serving a fairly niche audience, and unfortunately does get pushed down the list. The timing with the pending update is pretty shit, and I’d be annoyed if I were expecting it sooner — but there's only one of me, and only so many hours in a day.

@algertc Hows it going? Anything we can do to help?


And if anyone wants to vibe code a pull request and share a short screen recording of it working, I'd be glad to review and get it merged.

Pls post here first if u are someone who wants to try to do this. Will avoid issues later on if I can provide the current progress.
 
Last edited:
  • Wow
Reactions: iwanttosee
Sorry to hear about your situation!!! You've been a great asset to our little community, so we all wish you well.

I did vibe code two quick fixes to some issues, which I posted here awhile back:


I didn't want to bother you with it at the time, but since you asked, here you go!
 
I did vibe code two quick fixes to some issues, which I posted here awhile back:


I didn't want to bother you with it at the time, but since you asked, here you go!

This is a good catch on the logging ty. The middleware vulnerability is important, but there is a reason for the sidecar that was dropped. I have this sort of swapped for a different system in my code currently which will allow for other drop in things like adding another container to the compose to run a tiny model for semantic search.

Can’t quite merge this directly in with the commit it has, but I will use the logging fix. Thank you.
 
  • Like
Reactions: TheWaterbug
This is a good catch on the logging ty. The middleware vulnerability is important, but there is a reason for the sidecar that was dropped. I have this sort of swapped for a different system in my code currently which will allow for other drop in things like adding another container to the compose to run a tiny model for semantic search.

Can’t quite merge this directly in with the commit it has, but I will use the logging fix. Thank you.
It was all Claude! I was just watching. It was actually funny; Claude was soooo convinced that 100% of my performance issues were because of this logging thing, and then when it found the malware it actually put exclamation points on its response put its response in ALL CAPS:

The Node.js process is healthy per stats.ndjson — but 125 PIDs vs 23 is suspicious. Something is spawning many processes. Let me see what's actually running.
Ran: List all container processes sorted by CPU
STOP EVERYTHING — this is malware, not a memory leak.

PID 70 (/v0UDjR -c /b27pyQ -B) is running at 406% CPU and 2.4 GiB RAM with an obfuscated binary name at the filesystem root. The -c config -B flags are the exact invocation pattern of XMRig, a Monero cryptocurrency miner. The dozen zombie [let] and [BcXsGyx] processes are its spawn history.

The miner is what's been pegging your CPU and RAM, not the Node.js app. The pkill -f sh is the miner's anti-debugging self-defense — it kills shell processes to evict debuggers. We've been restarting a compromised container repeatedly.