Grafana Support enhancements

[hsscatlas]

Following up the discussion started here: #935 (reply in thread)

Last point: I had some trouble setting up time series on grafana using the timeplus/proton plugin. The same visualization works perfectly fine with the ClickHouse plugin, but on Proton it is impossible to have time series for multiple identifiers (say id, timestamp, max_temperature, for instance).

This enhancement would be really appreciated.

Cheers,

Hugo

[jovezhong]

Hi Hugo, thanks for creating the new thread. Can you share a bit more details, such as db schema, sample data, and SQL, ideally with screenshot of Grafana dashboard for ClickHouse and Timeplus, so that we can reproduce and enhance this.

[hsscatlas]

Sure. Find below two snapshots of what is expected to be similar graphs in Grafana.

The ClickHouse's "agg_well_data" is receiving the results from Proton. The query used in the MV is exactly the same I use in the Proton query builder below.
As you can see, with the ClickHouse connector, I can have different series based on the same data that is being produced by Proton. On Proton, I don't have the possibility to have a "Time Series".
As you can see from the sample data, the fact of having a Time Series produces a table pivot. That's exactly the feature I think the Proton plugin is missing.

Time Series graphs:

• ClickHouse
  

• Proton
  

Sample data:

• ClickHouse
  

• Proton
  

Hope it helps.

Cheers,

Hugo

[ye11ow]

Hi @hugo-catlas
You can just use the Transformations feature of Grafana to convert the result of Proton into the time series format.

1. Make sure the first column is the timestamp (e.g. select window_start, the rest of the columns from table)
2. Go to "Transformations" tab
3. Select "Prepare time series"
4. Select "Multi-frame time series" as the Format

This is what I got with my testing data
SQL: select window_start, avg(total_km) as km, cid from hop(car_live_data,3s, 3s) where (cid='c00006' or cid='c00001') group by cid, window_start

Please let me know if this works or not. Thanks!

[jovezhong]

Hi @hugo-catlas , feel free to let us know if you need more help. Happy to have a meeting to discuss more. You can join https://timeplus.com/slack for DM or private channel.

[jovezhong]

Hi @hugo-catlas , any chance to check answer from Calvin and let us know whether you have any more question.

[hsscatlas]

@jovezhong Apologies for my late reply.
Indeed, your instructions work like a charm. Sorry for having overlooked that.
Cheers,
Hugo

[jovezhong]

Glad to hear that. Thanks for marking the question is answered. We keep enhancing the plugin per user feedback like you. Yesterday there is a new version with improved nullable data type support. If you get more question or feedback, feel free to contact us. We also would like to invite you to share how you use Timeplus via our monthly webinar or blog.