Metrics, forecast, and outlier detection (autoML) automation

1

Hey there, new to the community. I’d like to start a discussion.

We are the team behind Loud ML (loudml.io). Several Grafana community members requested us to work on an integration of Loud ML inside Grafana. It’s possible, so we’d like to ask you a few questions:

What are the predictive functions you’d like to see embedded inside Grafana?

What are the metrics of interest that you would like to baseline for automated outlier detection? (AutoML)

What are the metrics of interest that you would like to forecast?

Would you be interested to help us and contribute to the Open Source development of a Grafana integration?

Looking forward to your feedback

Sébastien and the team.
Twitter: @loud_ml

6 Likes
2

anything new with that?

3

Hi Regel, is LoudML integrated with Grafana? Please advise.

4

Hello guys,

I’m working on a plugin with graph and datasource to use Loud ML to train and run ML models. Right now it is capable to create a model in 1-click and show predictions and anomalies with annotations.

I have recorder a small video about how it looks https://youtu.be/bxfU1N3ut70

It is still in a very dev stage, please see it on GitHub https://vsergeyev.github.io/loudml-grafana-app/

Regards,
Volodymyr

5

Regard @vsergeyev, can you make docker image with your plugin?

Thank you.

6

Hello @misho

Trying to build such image right now.

Regards,
V.

1 Like
7

Hello @misho,

It is new for me to make docker image with grafana.
I tried to build it locally from sources, not too much luck.

What I managed to get working is:

sudo docker run -d \
-p 3000:3000 \
--name=grafana \
-e "GF_INSTALL_PLUGINS=https://github.com/vsergeyev/loudml-grafana-app/raw/master/grafana-loudml-app-1.1.0.zip;grafana-loudml-app" \
grafana/grafana

This creates a docker container with vanilla Grafana and Loud ML app installed in plugins.
Please let me know if this is works for you.

V.

1 Like
8

I’m trying to run loudml in ubunut and this is the error I receive

sudo systemctl status loudmld.service

● loudmld.service - Loud ML daemon
   Loaded: loaded (/lib/systemd/system/loudmld.service; enabled; vendor preset: enabled)
  Drop-In: /etc/systemd/system/loudmld.service.d
           └─override.conf
   Active: failed (Result: exit-code) since Thu 2021-08-12 15:34:20 UTC; 4s ago
  Process: 6755 ExecStart=/usr/bin/loudmld (code=exited, status=1/FAILURE)
 Main PID: 6755 (code=exited, status=1/FAILURE)

Aug 12 15:34:20 seeqserver systemd[1]: loudmld.service: Service hold-off time over, scheduling restart.
Aug 12 15:34:20 seeqserver systemd[1]: loudmld.service: Scheduled restart job, restart counter is at 5.
Aug 12 15:34:20 seeqserver systemd[1]: Stopped Loud ML daemon.
Aug 12 15:34:20 seeqserver systemd[1]: loudmld.service: Start request repeated too quickly.
Aug 12 15:34:20 seeqserver systemd[1]: loudmld.service: Failed with result 'exit-code'.
Aug 12 15:34:20 seeqserver systemd[1]: Failed to start Loud ML daemon.
ubuntu@seeqserver:~$ sudo journalctl --unit loudmld
-- Logs begin at Mon 2021-07-12 18:04:56 UTC, end at Thu 2021-08-12 15:35:09 UTC. --
Aug 11 18:33:13 seeqserver systemd[1]: Started Loud ML daemon.
Aug 11 18:33:13 seeqserver loudmld[18068]: Traceback (most recent call last):
Aug 11 18:33:13 seeqserver loudmld[18068]:   File "/usr/bin/loudmld", line 5, in <module>
Aug 11 18:33:13 seeqserver loudmld[18068]:     from loudml.server import main
Aug 11 18:33:13 seeqserver loudmld[18068]: ModuleNotFoundError: No module named 'loudml'
Aug 11 18:33:13 seeqserver systemd[1]: loudmld.service: Main process exited, code=exited, status=1/FAILURE
Aug 11 18:33:13 seeqserver systemd[1]: loudmld.service: Failed with result 'exit-code'.
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Service hold-off time over, scheduling restart.
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Scheduled restart job, restart counter is at 1.
Aug 11 18:33:14 seeqserver systemd[1]: Stopped Loud ML daemon.
Aug 11 18:33:14 seeqserver systemd[1]: Started Loud ML daemon.
Aug 11 18:33:14 seeqserver loudmld[18091]: Traceback (most recent call last):
Aug 11 18:33:14 seeqserver loudmld[18091]:   File "/usr/bin/loudmld", line 5, in <module>
Aug 11 18:33:14 seeqserver loudmld[18091]:     from loudml.server import main
Aug 11 18:33:14 seeqserver loudmld[18091]: ModuleNotFoundError: No module named 'loudml'
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Main process exited, code=exited, status=1/FAILURE
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Failed with result 'exit-code'.
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Service hold-off time over, scheduling restart.
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Scheduled restart job, restart counter is at 2.
Aug 11 18:33:14 seeqserver systemd[1]: Stopped Loud ML daemon.
Aug 11 18:33:14 seeqserver systemd[1]: Started Loud ML daemon.
Aug 11 18:33:14 seeqserver loudmld[18112]: Traceback (most recent call last):
Aug 11 18:33:14 seeqserver loudmld[18112]:   File "/usr/bin/loudmld", line 5, in <module>
Aug 11 18:33:14 seeqserver loudmld[18112]:     from loudml.server import main
Aug 11 18:33:14 seeqserver loudmld[18112]: ModuleNotFoundError: No module named 'loudml'
Aug 11 18:33:14 seeqserver systemd[1]: loudmld.service: Main process exited, code=exited, status=1

Can anyone support?