feat(coderd/database): use template_usage_stats in GetTemplateInsightsByInterval query (#12667)

This PR updates the `GetTemplateInsightsByInterval` query to use rolled up `template_usage_stats` instead of raw agent and app stats.
This commit is contained in:
Mathias Fredriksson
2024-03-25 15:45:49 +02:00
committed by GitHub
parent 35d08434a9
commit 5f3be62c83
3 changed files with 126 additions and 177 deletions
+26 -55
View File
@@ -297,64 +297,35 @@ GROUP BY template_id, display_name, slug_or_port;
-- time, if end time is a partial interval, it will be included in the results and
-- that interval will be shorter than a full one. If there is no data for a selected
-- interval/template, it will be included in the results with 0 active users.
WITH ts AS (
SELECT
d::timestamptz AS from_,
CASE
WHEN (d::timestamptz + (@interval_days::int || ' day')::interval) <= @end_time::timestamptz
THEN (d::timestamptz + (@interval_days::int || ' day')::interval)
ELSE @end_time::timestamptz
END AS to_
FROM
-- Subtract 1 microsecond from end_time to avoid including the next interval in the results.
generate_series(@start_time::timestamptz, (@end_time::timestamptz) - '1 microsecond'::interval, (@interval_days::int || ' day')::interval) AS d
), unflattened_usage_by_interval AS (
-- We select data from both workspace agent stats and workspace app stats to
-- get a complete picture of usage. This matches how usage is calculated by
-- the combination of GetTemplateInsights and GetTemplateAppInsights. We use
-- a union all to avoid a costly distinct operation.
--
-- Note that one query must perform a left join so that all intervals are
-- present at least once.
SELECT
ts.*,
was.template_id,
was.user_id
FROM ts
LEFT JOIN workspace_agent_stats was ON (
was.created_at >= ts.from_
AND was.created_at < ts.to_
AND was.connection_count > 0
AND CASE WHEN COALESCE(array_length(@template_ids::uuid[], 1), 0) > 0 THEN was.template_id = ANY(@template_ids::uuid[]) ELSE TRUE END
WITH
ts AS (
SELECT
d::timestamptz AS from_,
CASE
WHEN (d::timestamptz + (@interval_days::int || ' day')::interval) <= @end_time::timestamptz
THEN (d::timestamptz + (@interval_days::int || ' day')::interval)
ELSE @end_time::timestamptz
END AS to_
FROM
-- Subtract 1 microsecond from end_time to avoid including the next interval in the results.
generate_series(@start_time::timestamptz, (@end_time::timestamptz) - '1 microsecond'::interval, (@interval_days::int || ' day')::interval) AS d
)
GROUP BY ts.from_, ts.to_, was.template_id, was.user_id
UNION ALL
SELECT
ts.*,
w.template_id,
was.user_id
FROM ts
JOIN workspace_app_stats was ON (
(was.session_started_at >= ts.from_ AND was.session_started_at < ts.to_)
OR (was.session_ended_at > ts.from_ AND was.session_ended_at < ts.to_)
OR (was.session_started_at < ts.from_ AND was.session_ended_at >= ts.to_)
)
JOIN workspaces w ON (
w.id = was.workspace_id
AND CASE WHEN COALESCE(array_length(@template_ids::uuid[], 1), 0) > 0 THEN w.template_id = ANY(@template_ids::uuid[]) ELSE TRUE END
)
GROUP BY ts.from_, ts.to_, w.template_id, was.user_id
)
SELECT
from_ AS start_time,
to_ AS end_time,
array_remove(array_agg(DISTINCT template_id), NULL)::uuid[] AS template_ids,
COUNT(DISTINCT user_id) AS active_users
FROM unflattened_usage_by_interval
GROUP BY from_, to_;
ts.from_ AS start_time,
ts.to_ AS end_time,
array_remove(array_agg(DISTINCT tus.template_id), NULL)::uuid[] AS template_ids,
COUNT(DISTINCT tus.user_id) AS active_users
FROM
ts
LEFT JOIN
template_usage_stats AS tus
ON
tus.start_time >= ts.from_
AND tus.end_time <= ts.to_
AND CASE WHEN COALESCE(array_length(@template_ids::uuid[], 1), 0) > 0 THEN tus.template_id = ANY(@template_ids::uuid[]) ELSE TRUE END
GROUP BY
ts.from_, ts.to_;
-- name: GetTemplateUsageStats :many
SELECT