[CI] Fix build

[nemesifier]

Fix the build on github actions, which is failing.

[nepython]

The build is failing because influxdb container might be taking longer than 15s to start on the CI.
@nemesisdesign, I think initially we had similar issue which we dealt with in #23.
I tried out the below steps locally to confirm if the changes introduced in #23 are still effective.

$ sudo service influxdb stop
$ ./tests/manage.py makemigrations --dry-run monitoring
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 1 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 2 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 3 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 4 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 5 out of 5).
Traceback (most recent call last):
# and the urllib3 exception follows

[nemesifier]

The build is failing because influxdb container might be taking longer than 15s to start on the CI.
@nemesisdesign, I think initially we had similar issue which we dealt with in #23.
I tried out the below steps locally to confirm if the changes introduced in #23 are still effective.

$ sudo service influxdb stop
$ ./tests/manage.py makemigrations --dry-run monitoring
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 1 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 2 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 3 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 4 out of 5).
Got error while connecting to timeseries database. Retrying again in 3 seconds (attempt n. 5 out of 5).
Traceback (most recent call last):
# and the urllib3 exception follows

I don't see this output in the build output: https://github.com/openwisp/openwisp-monitoring/runs/1738993241

@devkapilbansal can you look into it when possible?
Also, I didn't notice this, but why is the redis service declared also in the deploy step, is it needed there?