kylebrandt

Full Long Formats

Two new data types have been introduced with SQL expressions. timeseries-full-long and numeric-full-long. This is a row-oriented tabular format similar to long (labels are not used), but structured in a way that allows more lossless conversion of information between the labeled (wide and multi) formats.

So the use case for this format is when you need to be able to convert between labeled and tabular formats with more preservation of the information. Whereas, when converting back and forth with Long can create artifacts.

Properties shared by Full Long Formats

They are single frame formats.

kylebrandt

SQL Expressions Go MySQL Fork Notes

Why a Fork?

For dolthub/go-mysql-server, MySQL compatibility is a top priority. For Grafana, avoiding Cgo is a top priority. dolthub/go-mysql-server uses Cgo for regex by default. You can disable it with a build tag, but then people have to know about that build tag for things like go test (and Mac users have issues with the dependency, we tried and go Makefile and CI working, but then realized we would still have this bump for devs).

So this is a minimal fork which makes non-cgo the default.

References

kylebrandt

Have PR and Commit links in a table with GH datasource. Join two Github datasource queries, extracting the PR id from the commit message.

One could do things get labbels for commits this way to (the labels of the PR that the commit came from)

SELECT
  grafana_prs.title AS pr_title,

kylebrandt

func TestWhereOnNullNumber(t *testing.T) {
	f := &data.Frame{
		RefID: "a",
		Name:  "a",
		Fields: []*data.Field{
			//data.NewField("d", nil, []*float64{p(2.35), p(math.NaN())}),
			data.NewField("d", nil, []float64{2.35, math.NaN()}),
		},
	}

kylebrandt

SQL Expressions -- Errors, No Data and Dynamic Schema Query Responses

There are three general situations that SQL expressions does not handle gracefully that users should be aware of that are not bugs, but rather current limitations of the feature. Those are:

1. Error Responses
2. No Data Responses
3. Dynamic Schema responses (Ones where columns are labels change between executions of the query)

Why are there these limitations?

kylebrandt

{
  "__inputs": [
    {
      "name": "DS_MOCK",
      "label": "Mock",
      "description": "",
      "type": "datasource",
      "pluginId": "grafana-mock-datasource",
      "pluginName": "Mock"
    },

kylebrandt

Dataplane FAQ

Note: much of this is moved to grafana/dataplane#56

How does dataplane relate to Dataframes?

Dataplane adds a data typing layer on top of data frames. By analogy Dataplane Types are to dataframes as typescript is to javascript.

Specifically, a property is added to the dataframe to indicate its type with is made up of a kind and a format, for example "Timeseries Wide".

kylebrandt

HTTP/1.1 200 OK
Audit-Id: 5b82aaf1-4ba7-48bc-bb7b-94a7b89c6c5e
Cache-Control: no-store
Content-Type: application/json
X-Content-Type-Options: nosniff
X-Frame-Options: deny
X-Xss-Protection: 1; mode=block
Date: Fri, 19 Sep 2025 14:51:41 GMT
Content-Length: 902
Connection: close

kylebrandt

The queries

Panel: Main Sheet

SQL:

SELECT title, author, copyright, href, image, rating, series
FROM main

kylebrandt

/* Long 
WITH movie_genre AS (
  SELECT
    A.title,
    A.year,
    B.Genre AS genre
  FROM A
  JOIN B ON A.imdb_id = B.imdb_id
),
genre_totals AS (