Full ALPR Database System for Blue Iris!

I think the easiest and most cost-effective solution is just going to be some mini fans with some weatherproof vents like these:

1737344813695.png
1737344843550.png
 
Pushed a small update:
Fixed known plates null value
Fixed camera name filter matching
Occurrence count now tracked in an actual column

Your database page will have zero for all the counts to start after the migration. You can fill them all in by going to /backfill and clicking the button to count them. After doing this once, you should not need to go to this page again.

The live feed page is still counting them since this will probably confuse a few people. I will definitely put it in the notes for a full update release. Very important change though.
 
Pushed a small update:
Fixed known plates null value
Fixed camera name filter matching
Occurrence count now tracked in an actual column

Your database page will have zero for all the counts to start after the migration. You can fill them all in by going to /backfill and clicking the button to count them. After doing this once, you should not need to go to this page again.

The live feed page is still counting them since this will probably confuse a few people. I will definitely put it in the notes for a full update release. Very important change though.

Updated manually on unRAID (running migrations.sql manually) and all seems to be fine. Made it count the occurrences and it also completed successfully. Thank you!
 
I do not know exactly what to do. I just ran across this a few months ago and thought of it when I read your thread.

Maybe this will help?


The POE step down to 5V might be generating a lot of heat, maybe you can do the step down outside of your enclosure to avoid complex cooling for your Pi and SDR? Maybe something like this 5V 1.8A Isolated Output PoE Module - For Raspberry Pi 3 B+ or 4
Maybe also consider using 23 AWG cat6 to limit the voltage drop if you are doing longer runs?
 
There was another docker update yesterday/today and the dashboard and live feed is empty at me following the update. No new plates are getting recorded either.
Does anyone has an idea how to downgrade the docker (I mean what shall I pull instead of :latest)?
 
Last edited:
There was another docker update yesterday/today and the dashboard and live feed is empty at me following the update. No new plates are getting recorded either.
Does anyone has an idea how to downgrade the docker (I mean what shall I pull instead of :latest)?
Did you manually update the two below files before pulling the latest?

 
The schema.sql file is only used during the initial database creation, right? It wouldn't make sense to update it between database updates.
The migration.sql file didn’t change between the two Docker updates yesterday. You might have seen my post just before your issue above, where I mentioned that the first update, which included the occurrences change, worked fine for me.
However, a second update came after that, which broke my installation. Did you also get this second update?
 
The schema.sql file is only used during the initial database creation, right? It wouldn't make sense to update it between database updates.
The migration.sql file didn’t change between the two Docker updates yesterday. You might have seen my post just before your issue above, where I mentioned that the first update, which included the occurrences change, worked fine for me.
However, a second update came after that, which broke my installation. Did you also get this second update?
Both files were changes with the last update

1737461061562.png
 
I know that both files have been updated, but what I'm saying is that the update to schema.sql doesn't matter since that file is only used during fresh installations when the database is initially created. After that, it’s no longer used—not even for database updates.
It's the migrations.sql file that ensures the existing database is updated with new tables and columns.
Am I misunderstanding this?

...and also there was two updates in the last 24 hours. Have you got the second one? (I actually just tried a vanilla installation and that's failing as well).
I am getting the following errors in the log:
Code:
1/21/2025, 2:41:26 PM [ERROR] Error fetching plate reads: error: column pr.image_path does not exist
1/21/2025, 2:41:30 PM [INFO] GET /live_feed
1/21/2025, 2:41:30 PM [INFO] POST /api/verify-session
1/21/2025, 2:41:30 PM [INFO] Fetching latest plate reads
1/21/2025, 2:41:30 PM [INFO] Fetching tags

And also:
Code:
1/21/2025, 2:56:24 PM [ERROR] Error fetching dashboard metrics: error: column "thumbnail_path" does not exist

It seems I’ve received a version that already includes the new image handling rewrite, but it’s not yet complete?! Am I the only one experiencing this?
Could this somehow be related to the fact that I’m pulling alpr-dashboard from within unRAID?
I am confused...
 
There was another docker update yesterday/today and the dashboard and live feed is empty at me following the update. No new plates are getting recorded either.
Does anyone has an idea how to downgrade the docker (I mean what shall I pull instead of :latest)?

It seems I’ve received a version that already includes the new image handling rewrite, but it’s not yet complete?!
This is exactly what happened. Sorry about that! I pushed an image to test myself on another computer and didn't push the code to github.


The schema.sql file is only used during the initial database creation, right?
This is correct. The reason I update it is just for completeness. You do not need to have it in order to update.


I just pushed it and merged some other things too. Fully on new image storage now and the database got so much faster. Changes were made to the compose file and the migrations file. You will also need to create a new directory called "storage" in the same place as your config and auth folders. This is where the JPEGs will be stored from now on. You will notice a little yellow question mark in the bottom corner with a temporary explanation. There are links to tools that will automatically convert your existing data in that page.
 
This is exactly what happened. Sorry about that! I pushed an image to test myself on another computer and didn't push the code to github.



This is correct. The reason I update it is just for completeness. You do not need to have it in order to update.


I just pushed it and merged some other things too. Fully on new image storage now and the database got so much faster. Changes were made to the compose file and the migrations file. You will also need to create a new directory called "storage" in the same place as your config and auth folders. This is where the JPEGs will be stored from now on. You will notice a little yellow question mark in the bottom corner with a temporary explanation. There are links to tools that will automatically convert your existing data in that page.
The migration is failing

1737487626612.png

1737487598663.png
 
  • Like
Reactions: Vettester
I think this is because you have such large databases and it's maxing out the memory or the max query size. It worked fine for me. I tried to set it up to do it in batches and have an image building now for that. Not sure if this way will work, but if someone could try it again and share what happens, that would be very helpful.