Eugene Klimov

Could you try to replace your clickhouse query format from `time series` to `table`? (look drop-down menu on bottom) or try to replace your query and apply $columns macro ```sql...

Let's try remove `intDIv` and double group by something like that ```sql SELECT t, groupArray((cat, errors)) AS groupArr FROM ( SELECT 1 AS t, concat('Category ', toString(cat)) as cat, count(idx)...

### I ask third time!!! Did you try to replace your clickhouse query format from `time series` to `table`? (look drop-down menu on bottom)? 

ok ;) clear, thanks a lot for reporting, i will try to figure wrong behavior ASAP

we need to figure out and ensure could grafana core pass some metadata in headers we easily can add required metadata in query body with comment

@hlf2016 sorry for late response could you clarify what mean "clickhouse 1.1.2" ? this is clickhouge-grafana plugin version? or clickhouse-server version? could you share screenshot from `/plugins` endpoint for your...

@hyberdk thanks a lot for your reporting I will try add this functionality to next release

@kotchuprik big thanks for a workaround unfortunately, I don't have any idea how to properly recognize when we shall use JSON.parse and when is not cause ClickHouse database doesn't have...

@kotchuprik yep, but this solution will decrease performance when data amount will grow

no, golang part of plugin use only for alerts maybe it can be handled by the "Transform" feature on grafana on the browser side? https://grafana.com/blog/2020/06/11/new-in-grafana-7.0-data-transformations-for-all-visualizations-that-support-queries/