You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
Timeseries graphs using the redshift $__timeGroup macro are incorrect as groups with no data are not filled in. For example, the query:
SELECT
$__timeGroup(timestamp, '10m'),
count($__column)
FROM
$__schema.$__table
WHERE
$__timeFilter(timestamp)
and $__column = '$Seller'
GROUP BY
time
ORDER BY
time
Results in:
time
count
2025-01-15 14:00
1
2025-01-15 14:10
1
2025-01-15 14:30
1
2025-01-15 14:40
1
2025-01-15 15:00
1
2025-01-15 15:20
1
2025-01-15 15:30
1
2025-01-15 15:40
2
2025-01-15 16:10
2
2025-01-15 16:40
1
2025-01-15 16:50
3
2025-01-15 17:00
1
2025-01-16 05:00
1
2025-01-16 06:00
2
Notice the gap at 14:20, 14:50 etc. As there is no row, the grafana Fill options don't do anything. This results in timeseries graphs being incorrect
Describe the solution you'd like
Rows should be generated and the value returned as NULL for missing entries in the time group.
Describe alternatives you've considered
Creating a data source which is just timestamps to left join
Is your feature request related to a problem? Please describe.
Timeseries graphs using the redshift
$__timeGroup
macro are incorrect as groups with no data are not filled in. For example, the query:Results in:
Notice the gap at 14:20, 14:50 etc. As there is no row, the grafana Fill options don't do anything. This results in timeseries graphs being incorrect
Describe the solution you'd like
Rows should be generated and the value returned as
NULL
for missing entries in the time group.Describe alternatives you've considered
Creating a data source which is just timestamps to
left join
Additional context
Postgres has support for this feature: https://grafana.com/docs/grafana/latest/datasources/postgres/#macros
The text was updated successfully, but these errors were encountered: