Found a Real Hole in a major system

[rohithsiddabattuni-creator]

Describe the bug

I've analyzed the Redis pool exhaustion on high-traffic bursts. The synchronous increment in the downloads view is the bottleneck. I have a PR ready using django_redis async connection to decouple the stats from the request thread

To Reproduce

High concurrency on /downloads/ during release peaks causes redis.exceptions.ConnectionError

Expected behavior

Redis operations should be asynchronous and decoupled from the main request/response cycle using on_commit hooks or Celery tasks to prevent pool exhaustion when workers=2."

URL to the issue

No response

Screenshots

"![SCREENSHOT_DESCRIPTION](SCREENSHOT_LINK.png)"

Browsers

Chrome

Operating System

Windows

Browser Version

No response

Relevant log output

Paste the redis.exceptions.ConnectionError: Connection pool exhausted error.

Additional context

No response

[linear-code]

PYORG-2981

[rohithsiddabattuni-creator]

Summary of the Linear Bottleneck:
The current Redis connection pool exhaustion is a result of synchronous blocking during high-concurrency download spikes. With workers = 2, the pool is starved whenever Redis latency spikes, blocking the entire request/response cycle.
The Fix (Plug):
We need to decouple the download statistics from the main thread. By using transaction.on_commit, we ensure the Redis increment only happens after the response is cleared, preventing the pool from holding up the next worker request

[rohithsiddabattuni-creator]

Describe the bug

I've analyzed the Redis pool exhaustion on high-traffic bursts. The synchronous increment in the downloads view is the bottleneck. I have a PR ready using django_redis async connection to decouple the stats from the request thread

To Reproduce

High concurrency on /downloads/ during release peaks causes redis.exceptions.ConnectionError

Expected behavior

Redis operations should be asynchronous and decoupled from the main request/response cycle using on_commit hooks or Celery tasks to prevent pool exhaustion when workers=2."

URL to the issue

No response

Screenshots

""

Browsers

Chrome

Operating System

Windows

Browser Version

No response

Relevant log output

Paste the redis.exceptions.ConnectionError: Connection pool exhausted error.

Additional context

No response