Visualize MySQL EXPLAIN ANALYZE FORMAT=JSON output as interactive SVG charts. Four output types, one parser, no external dependencies.

Inspired by Brendan Gregg's FlameGraph and Tanel Poder's SQL Plan FlameGraphs.

Every view includes a Query Analysis panel below the chart with optimizer features detected, warnings (full table scans, hash joins, BNL join buffers, temp tables, filesorts), and tuning suggestions.

Note: Interactive features (zoom, search, tooltips) require the SVG to be opened from an HTML wrapper or GitHub Pages — not from a raw GitHub URL, which blocks inline scripts.

All demos index

• Python 3.7+ — no extra packages
• MySQL 8.4+ with explain_json_format_version = 2

Enable the required JSON format:

SET explain_json_format_version = 2;
-- or permanently in my.cnf:
-- [mysqld]
-- explain_json_format_version = 2

EXPLAIN ANALYZE FORMAT=JSON
SELECT u.name, COUNT(o.id)
FROM users u
JOIN orders o ON u.id = o.user_id
WHERE u.country = 'US'
GROUP BY u.id;

Save to a file:

mysql -u user -p mydb -N -e "EXPLAIN ANALYZE FORMAT=JSON SELECT ..." > explain.json

Or pipe directly (stdin supported):

mysql -u user -p mydb -N -e "EXPLAIN ANALYZE FORMAT=JSON SELECT ..." | python3 -m myflames > query.svg

# Flame graph (default)
python3 -m myflames explain.json > query.svg

# Bar chart — slowest operations first
python3 -m myflames --type bargraph explain.json > query-bar.svg

# Treemap — area proportional to total time
python3 -m myflames --type treemap explain.json > query-treemap.svg

# Diagram — Visual Explain-style flow
python3 -m myflames --type diagram explain.json > query-diagram.svg

open query.svg        # macOS
xdg-open query.svg    # Linux
start query.svg       # Windows

git clone https://github.com/vgrippa/myflames.git
cd myflames

Run from the repo root. No pip install needed.

python3 -m myflames [--type TYPE] [options] explain.json > output.svg

Text in the details strip is always selectable — you can copy/paste it freely.

Every output type includes a panel below the chart with:

• How to read — view-specific guide
• Optimizer features — e.g. index_condition_pushdown=on, batched_key_access
• Warnings — issues that affect performance:
  • Full table scans (with row count)
  • Hash joins (with estimated build phase size)
  • Block Nested-Loop (BNL) join buffers
  • Temp tables / Materialize operations
  • Filesorts
• Suggestions — concrete tuning actions (add indexes, increase join_buffer_size, enable hash join, etc.)

Warnings also show which node label in the chart they refer to, so you can find the slow operation quickly.

• Width = time — wider frames consumed more time (including children)
• Bottom = root — the query entry point; table accesses are at the top
• Self-time — the visible "tip" of each frame is time spent in that operation alone

Frame labels follow Tanel Poder's format:

OPERATION [table.index] starts=X rows=Y

Example: INDEX LOOKUP [orders.idx_user] starts=1000 rows=5 → This lookup ran 1000 times in a nested loop, returning 5 rows each time.

• Sorted slowest first by self-time (time in that operation, not counting children)
• Percentage shows each operation's share of total query time

• Area = total time (including all descendants)
• Nested rectangles show the parent/child relationship
• Color intensity indicates relative cost

• Left-to-right execution flow (table accesses → join → result)
• Darker color = more time spent at that step
• Arrows show row flow with estimated row counts
• Diamonds represent nested-loop join decision points

All views automatically switch units based on total query time:

python3 -m myflames --title "Before" before.json > before.svg
python3 -m myflames --title "After"  after.json  > after.svg

python3 -m myflames --type bargraph --title "Before" before.json > before-bar.svg
python3 -m myflames --type bargraph --title "After"  after.json  > after-bar.svg

python3 stackcollapse_mysql_explain_json.py explain.json > stacks.txt

Useful for feeding into other FlameGraph-compatible tools.

"No module named 'myflames'" Run from the repo root, or use the wrapper scripts directly: python3 mysql_explain.py.

Empty output or parse error Make sure you're using EXPLAIN ANALYZE FORMAT=JSON (not just EXPLAIN FORMAT=JSON). The ANALYZE keyword is required for timing data.

Interactive features not working Open the .html wrapper file instead of the raw .svg. Browsers block inline scripts in SVGs loaded from raw.githubusercontent.com. Local file:// access works fine.

Graphviz diagram falls back to built-in Install Graphviz and ensure dot is on your PATH, then use --diagram-engine graphviz.

• Brendan Gregg — FlameGraph implementation (pure-Python port in myflames/flamegraph.py)
• Tanel Poder — SQL Plan FlameGraph concept and label format

Extends Brendan Gregg's FlameGraph project. See docs/cddl1.txt (CDDL 1.0).