teslamate
teslamate copied to clipboard
teslamate-org
Update statistics.json: timezone issues addressed
Deploy Preview for teslamate ready!
| Name | Link |
|---|---|
| Latest commit | ffd8ba7a59e9eb301fb66294027820818dc98f63 |
| Latest deploy log | https://app.netlify.com/sites/teslamate/deploys/6627b3ab2dbbc5000834ae99 |
| Deploy Preview | https://deploy-preview-3862--teslamate.netlify.app |
| Preview on mobile | Toggle QR Code...Use your smartphone camera to open QR code link. |
To edit notification comments on pull requests, go to your Netlify site configuration.
![netlify[bot] avatar](https://avatars.githubusercontent.com/in/13473?v=4 "Published by a pub.dev verified publisher"){kind=small}
To test this changes, we can update our docker-compose.yml as follows (see https://docs.teslamate.org/docs/development#testing-with-our-ci-which-builds-the-docker-images-automatically-per-pr)
grafana:
# image: teslamate/grafana:latest
image: ghcr.io/teslamate-org/teslamate/grafana:pr-3862
do a docker compose pull followed by a docker compose up -d afterwards
To test this changes, we can update our
docker-compose.ymlas follows (see https://docs.teslamate.org/docs/development#testing-with-our-ci-which-builds-the-docker-images-automatically-per-pr)grafana: # image: teslamate/grafana:latest image: ghcr.io/teslamate-org/teslamate/grafana:pr-3862do a
docker compose pullfollowed by adocker compose up -dafterwards
Going to ask a dumb question here (forgive me). How would you remove these 'test' images once you are done testing? Or does it not retain them somehow? Mainly thinking space on my Pi.
Wondering if "timezone" in all of our dashboard JSONs should be "browser". https://github.com/search?q=repo%3Ateslamate-org%2Fteslamate%20timezone&type=code
This does not work for me. Or I am not pulling the image correctly 🤷🏻♂️ I continue to see the timezone dropdown. Looking through the JSON, I do not see the iteration to verify it was upped. Followed the guidance above:
grafana:
# image: teslamate/grafana:latest
image: ghcr.io/teslamate-org/teslamate/grafana:pr-3862
But still not working. Tried to prune all of my docker images with docker image prune -a and docker container prune to no luck.
How would you remove these 'test' images once you are done testing? Or does it not retain them somehow? Mainly thinking space on my Pi.
Good point. There are several ways, but if you want to remove dangling images, this would be safest imo:
docker image prune -a
WARNING! This will remove all images without at least one container associated to them.
see https://docs.docker.com/reference/cli/docker/image/prune/
This needs to be tested still... Edge cases are drives and charges between midnight UTC and midnight local time.
Wondering if "timezone" in all of our dashboard JSONs should be "browser". https://github.com/search?q=repo%3Ateslamate-org%2Fteslamate%20timezone&type=code
AFAICS one is browser, the others are empty, probably also meaning browser.
I am a bit struggling here... I have two charges: in the database:
id | start_date
-----+-------------------------
576 | 2024-01-20 21:37:50.375
577 | 2024-01-20 23:01:09.237
(2 rows)
Since it is UTC and my local time is Europe/Berlin, the second was indeed on 2024-01-21 00:01 and should show up for the 2024-01-21, but it show up for 2024-01-20.
Relevant query:
WITH data AS (
SELECT
charging_processes.*,
date_trunc('$period', TIMEZONE('$__timezone', start_date)) as local_period
FROM charging_processes)
SELECT
EXTRACT(EPOCH FROM date_trunc('$period', TIMEZONE('$__timezone', local_period)))*1000 AS date_from,
EXTRACT(EPOCH FROM date_trunc('$period', TIMEZONE('$__timezone', local_period) + ('1 ' || '$period')::INTERVAL))*1000 AS date_to,
CASE '$period'
WHEN 'month' THEN to_char(local_period, 'YYYY Month')
WHEN 'year' THEN to_char(local_period, 'YYYY')
WHEN 'week' THEN 'week ' || to_char(local_period, 'WW') || ' starting ' || to_char(local_period, 'YYYY-MM-DD')
ELSE to_char(local_period, 'YYYY-MM-DD')
END AS display,
local_period AS date,
sum(greatest(charge_energy_added,charge_energy_used)) AS sum_consumption_kwh,
sum(greatest(charge_energy_added,charge_energy_used)) / count(*) AS avg_consumption_kwh,
sum(cost) AS cost_charges,
count(*) AS cnt_charges
FROM data WHERE
car_id = $car_id AND
$__timeFilter(start_date) AND
(charge_energy_added IS NULL OR charge_energy_added > 0.1)
GROUP BY date
ORDER BY date
I tried to apply the timezone correction to the other, but it did not help. Any thoughts?
This does not work for me. Or I am not pulling the image correctly 🤷🏻♂️ I continue to see the timezone dropdown. Looking through the JSON, I do not see the iteration to verify it was upped.
Same here. But the images have been produced...
This does not work for me. Or I am not pulling the image correctly 🤷🏻♂️ I continue to see the timezone dropdown. Looking through the JSON, I do not see the iteration to verify it was upped.
Same here. But the images have been produced...
@JakobLichterfeld is there something we are missing? Does it go in Grafana or TeslaMate service?
@JakobLichterfeld is there something we are missing? Does it go in Grafana or TeslaMate service?
Not really. It must be included in the grafana image. It on my side as well, and your changes seems to be not correctly reflected in the Grafana image. Currently, I'm out of time to dig deeper. (https://github.com/teslamate-org/teslamate/issues/3857#issuecomment-2068640395)
I'd like some more testing from users with significant timezone difference to UTC, e.g. UTC+-5 or more...
I'd like some more testing from users with significant timezone difference to UTC, e.g. UTC+-5 or more...
UTC -5 (or 6 depending on the time of year) hours and just tested.
- Timezone dropdown is removed.
- Verified that it starts at 12AM in month and day view.
- Month view looks correct.
- Clicking on a day, the timeframe is midnight to midnight of the selected day.
One item I did find, testing a use-case from @towster - it seems the Drives dashboard filters out (either) zero duration or zero distance drives. I saw this when comparing the filtered Drives dashboard versus the same day Trip dashboard links off the Statistic page. This might be by design.
I am approving and merging. Linking all issues such that they are closed when merged.
Great work @DrMichael and thank you 😄
One item I did find, testing a use-case from @towster - it seems the Drives dashboard filters out (either) zero duration or zero distance drives. I saw this when comparing the filtered Drives dashboard versus the same day Trip dashboard links off the Statistic page. This might be by design.
Right, it filters out also drives with duration_min=0 although there is distance>0. I proposed to show those also on the Drives dashboard in https://github.com/teslamate-org/teslamate/pull/3867
One item I did find, testing a use-case from @towster - it seems the Drives dashboard filters out (either) zero duration or zero distance drives. I saw this when comparing the filtered Drives dashboard versus the same day Trip dashboard links off the Statistic page. This might be by design.
Right, it filters out also drives with duration_min=0 although there is distance>0. I proposed to show those also on the Drives dashboard in #3867
Found this PR about five minutes after making the comment above 🤦🏼 thanks for the find.
@DrMichael Seems like there is an issue with this dashboard now:
teslamate_grafana | logger=provisioning.dashboard type=file name=teslamate t=2024-05-20T13:33:21.500815425Z level=error msg="failed to save dashboard" file=/dashboards/statistics.json error="could not resolve dashboards:uid:1EZnXszMl: Dashboard not found"
Where does that come from? The build? Not sure, what that uid issue means...
(After a pull for edge I am still on 1.28.6-dev, might be an issue with my edge installation again... latest gives me 1.29.1. But no issue with Statistics...)
@DrMichael Seems like there is an issue with this dashboard now:
Where does that come from?
Seems to be a dangling container issue on my end, grafana was on 1.29.1 and teslamate was on 1.28.5 hence the mismatch. Sorry for the inconvenience
@DrMichael I am still seeing the timezone in 1.29.1 and it defaults to Berlin. Is anyone else seeing this? @JakobLichterfeld @jheredianet
Hmmm, looks good for me. You are running grafana:latest or grafana:edge?
I had a terrible time some weeks ago to get rid of old images. Perhaps try to purge all images of grafana and repull it.
Not on edge. Running latest across all images. Just updated to 1.29.1 last night. Part of my upgrade script is a prune after pulling the latest image.
@DrMichael I am still seeing the timezone in 1.29.1 and it defaults to Berlin. Is anyone else seeing this? @JakobLichterfeld @jheredianet
Can confirm Timezone setting on Statistics dashboard is still there with 1.29.1
Hmmm, that is wierd. I removed all teslamate and grafana images, did a fesh pull. And have the new version...
Is there something in the dashboard JSON I can verify to ensure it is in-fact the latest image?
The new one uses the new $__timezone variable. The old one probably has Europe/Berlin in it.
Looks like it is still there. This at least from the live Grafana interface. Not sure how to dive into the code in the Docker container specifically (if necessary).
The screenshots are from github master? You can check in the image. It is in /dashboards.