Blue Iris occasionally does not add its own files to the data base

[ManUnited10]

Hi all.

I have a blue iris on a Proxmox windows 10 vm with 20+ cameras.

In the past few months it is not behaving well. Occasionally, it somehow forgets to put the video files in "New" folder in its own database and I come up with nothing for a few days when I look back in the timeline or calendar.
The files are already there, but somehow BI thinks they don't belong to it! As a consequence, the drive grows and BI goes mad about the drive usage (it shows a lot of "other files" in grey). As a solution I have to rebuild the database and it's all fine again until it happens again.

Last time, I thought maybe it's some permission issue or something which BI cannot delete or move the files from drive to drive, but no things were good.
Anyway, I even formatted my SSD and rebuilt the database, still the issue is there... Yesterday I rebuild the database and now it's fine.

Any help would be appreciated.

Spec:
Windows 10 under Proxmox vm.
CPU i7-6850K, 4 cores passthrough to vm
RAM 128GB, 16GB passthrough to vm
Main OS drive is 64GB (drive C)
one SSD 1TB is directly passthrough to the vm (Drive G)
one HDD 8TB is directly passthrough to the vm (drive H)
Blue Iris is v5.7.9.4
The machine is fully disconnected from the internet and lives in a seperate VLAN with cameras.

Directory management in BI:
DB is on drive G (1TB SSD).
NEW is on drive G (1TB SSD), Limit Size is 500GB, Move to folder STORED
Alerts is on drive G (1TB SSD), Limit Size is 50GB, Limit clip age 30days, Delete
STORED is on drive H (8TB HDD), Limit Size is 7000GB, Delete
AUX1 -> not used
AUX2 (which is H:\BlueIris\faces) on drive H (8TB HDD), Limit Size 1 GB, Limit clip age 70days, Delete
The rest of AUX 3,4, ... are unused

The windows vm is not being used for anything else - just BI. There is no other files on drive G (1TB SSD). Only less than 1GB other files on drive H (8GB HDD).

 

[bp2008]

I've had issues like that before but not recently (Blue Iris's clip database has historically been a real mess). Since you're running a pretty old Blue Iris version I would suggest updating to latest. However doing that will require an active support plan.

Anything showing in Blue Iris's log that could indicate a failure to write a clip or database entry? If you don't have BI's logging turned on, I recommend you turn it on. It is rarely useful but takes up very little space typically so it is worth having.

Another thought. If drive G is filling up, that could theoretically lead to database write failure. Though BI does always try to keep 1 GB free on the disk I think, I don't know if that is enough to handle all possible database operations. You might try moving the database to drive C. That may require enlarging drive C but that is pretty straightforward with Proxmox. Just one command in the proxmox interface and then resize the partition in windows disk management. I think you might even be able to do it without shutting down windows. Or you could attach a whole new virtual disk -- thin provisioned -- just to dedicate to the clip database.

 

[ManUnited10]

Thank you for a quick reply.

Drive G is 1TB and I am assigning half of it i.e. 500GB to NEW and 50GB to Alerts, which means on paper 450GB is free. DB at the moment is less than 5GB which again on paper should be alright. I even put an automation on my home-assistant to let me know if the drives are filling up beyond their threshold points.

I can move the DB back to its default location drive C, but the whole reason that I did it was to easily move BI to another vm or another physical machine since everything lives in two physical disks (which are fully passthrough). Having said all, I can try that and move DB to drive C, which has enough free space - but I'm not sure that would do anything useful.

The logs are there in C:\Blueiris\logs (unless there's another one?). Each file is has a year and month name on it e.g. 202505.txt which I assume they hold a month of events. Last time I looked at them I found nothing in particular interesting. Ironically some of them are empty (0 bytes) and some are like a few hundred megabytes. Should I look for any special event?

Maybe as a last trial and error, I update BI. Generally I'm hesitant to update my home lab stuff unless I feel I am missing something big. Otherwise, if it works don't touch it!

 

[bp2008]

Having said all, I can try that and move DB to drive C, which has enough free space - but I'm not sure that would do anything useful.

Nah, I don't think it would help since you've confirmed there is lots of free space on the G drive.

I believe BI logs something every 5 minutes when doing file move or delete. If you can track down the exact place in the log where a database screw-up began, maybe there will be other clues there. For example if the size of tracked files goes down without BI reporting any file moves or deletions, that would indicate that database corruption occurred and any nearby log lines could be relevant, especially if they indicate that BI had just restarted after an unexpected shutdown.

It is also always a good idea to search log files for the word "error".

I don't blame you for the "don't fix it if it ain't broke" policy. I rarely update most of the BI installations I manage but if I have DB problems the first thing I do is update it to latest