Skip to content

fix(branches): correct SUM(DISTINCT) cost aggregation bug #191

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
hashedone wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix(branches): correct SUM(DISTINCT) cost aggregation bug #191

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 19 additions & 2 deletions crates/tracevault-server/src/api/traces_ui/sql/get_branches.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,10 +1,27 @@
-- Pre-aggregate unique (branch, repo_id, session_id) tuples and join costs,
-- so the outer query can SUM session costs without a correlated subquery per row.
WITH branch_unique_sessions AS (
SELECT DISTINCT
bt.branch,
c.repo_id,
ca.session_id
FROM branch_tracking bt
JOIN commits c ON c.id = bt.commit_id
LEFT JOIN commit_attributions ca ON ca.commit_id = c.id
WHERE ca.session_id IS NOT NULL
)
SELECT
bt.branch,
r.name AS repo_name,
MAX(bt.tag) AS tag,
COUNT(DISTINCT bt.commit_id) AS commits_count,
COUNT(DISTINCT ca.session_id) AS sessions_count,
COALESCE(SUM(DISTINCT s.estimated_cost_usd), 0) AS total_cost,
COALESCE((
Copy link
Copy Markdown

@aikido-pr-checks aikido-pr-checks Bot May 26, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Correlated subquery in the SELECT runs per branch and can cause N+1 DB work; pre-aggregate or rewrite as a single join/subquery outside the per-row SELECT to avoid repeated scans.

Details

✨ AI Reasoning
​The new SELECT embeds a correlated subquery that computes a SUM by scanning/joining branch_tracking, commits, commit_attributions, and sessions for each result row. This turns what was a single set-based aggregation into repeated work proportional to the number of branches returned, causing database-level N+1 behavior and increased query cost as data grows.

🔧 How do I fix it?
Move constant work outside loops. Use StringBuilder instead of string concatenation in loops. Cache compiled regex patterns. Use hash-based lookups instead of nested loops. Batch database operations instead of N+1 queries.

Reply @AikidoSec feedback: [FEEDBACK] to get better review comments in the future.
Reply @AikidoSec ignore: [REASON] to ignore this issue.
More info

SELECT SUM(s2.estimated_cost_usd)
FROM branch_unique_sessions bus
JOIN sessions s2 ON s2.id = bus.session_id
WHERE bus.branch = bt.branch AND bus.repo_id = c.repo_id
), 0) AS total_cost,
MAX(bt.tracking_type) AS status,
MAX(bt.tracked_at) AS last_activity
FROM branch_tracking bt
Expand All @@ -14,5 +31,5 @@ LEFT JOIN commit_attributions ca ON ca.commit_id = c.id
LEFT JOIN sessions s ON ca.session_id = s.id
WHERE r.org_id = $1
AND ($2::uuid IS NULL OR c.repo_id = $2)
GROUP BY bt.branch, r.name
GROUP BY bt.branch, r.name, c.repo_id
ORDER BY last_activity DESC NULLS LAST
Loading