Closed repomanz closed 3 years ago
example of mlapi add user script
documentation:
Running
Before you run, you need to create at least one user. Use python3 mlapi_adduser.py for that
command / error:
jonjames@vmubntml:/var/lib/zmeventnotification/mlapi$ sudo python3 mlapi_adduser.py Traceback (most recent call last): File "mlapi_adduser.py", line 8, in <module> db = Database.Database() File "/var/lib/zmeventnotification/mlapi/modules/db.py", line 12, in __init__ db = g.config['db_path']+'/db.json' KeyError: 'db_path'
content of mlapiconfig.ini:
db_path=./db
directory list:
jonjames@vmubntml:/var/lib/zmeventnotification/mlapi/db$ ls -l total 0 -rw-r--r-- 1 www-data www-data 0 Jan 9 17:12 empty
Hi, I am coming to MLAPI for the same reason as you and have come across the same issue when I run
python3 mlapi_adduser.py
did you manage to resolve this?
nope - I was hoping dev was going to answer but they haven't yet. My gut feel is that MLAPI isn't currently architectured to be stand alone.
I've just fixed the bug related to mlapi_adduser.py. As for the rest, the documentation is pretty straightfoward. mlapi is architected to be standalone from ZM. I did not respond because you closed it.
Hi @pliablepixels thanks for responding. I'll give this a go; thanks for fixing and clarifying.
Thanks, that’s fixed it. Now to try and figure out how to get the event server (running in dlandon’s docker) to connect to mlapi...
Hello -
I have been using dlandon's zoneminder container which up until recently included ES/ML options. I had these working well but he's decided to drop hooks / ml from his container but is keeping ES. Therefore I need to stand up a remote ML server to work along side his container so I can continue to do object detection.
I have a fresh Ubuntu 20.04 server stood up but what's not 100% clear to me is what specifically is required to standup a remote machine learning host. I have gone through your es/ms installation documentation as well as your mlapi documentation.
I believe I do require the mlapi project you have here (to handle the remote side of this) but looking through it, like for example the adding user part, it's looking for a db which doesn't natively exist with just mlapi. Therefore I assume that I need to do some basic installation of es/ml from your zmeventnotification project.
I have also looked at the zmeventnotification project, specifically around ES/ML installation and know the install script there will prompt me if I want things like ES/hooks/ml.
So, could you let me know a basic outline of what is required to stand up a remote ML server based on your two projects? Would love to get this going again. It was working great in his container but now I need to stand up a remote ml server to continue my object detection. The ml_sequence stuff was working great!