Skip to content

cloud-dep-inventory-refresh.zsh#24

Closed
benjermanjoel wants to merge 3 commits into
mainfrom
cloud-dp-inventory-refresh
Closed

cloud-dep-inventory-refresh.zsh#24
benjermanjoel wants to merge 3 commits into
mainfrom
cloud-dp-inventory-refresh

Conversation

@benjermanjoel

Copy link
Copy Markdown

Script Expectations

✅ Manager-approved

✅ Code Quality

✅ Security

✅ Testing

✅ Documentation

✅ Legal Requirements

@andrew-ladd

Copy link
Copy Markdown

After you've discovered N-1 pods, the chance of 10 consecutive probes all landing on already-seen pods (missing the last one) is:

• 2 pods → (1/2)^10 ≈ 0.1% — negligible
• 4 pods → (3/4)^10 ≈ 5.6% — non-trivial (~1 in 18 runs leaves a pod untouched)
• 8 pods → (7/8)^10 ≈ 26% — likely to miss at least one

The termination condition is 10 consecutive no-new-cookie probes — so as cluster size grows the risk of stopping too early increases significantly. Bumping POD_DISCOVERY_TRIES to 30+ for Premium instances would help, or switching to a fixed total probe count (e.g. 50 probes, collect all unique cookies seen) removes the statistical risk entirely.

@andrew-ladd andrew-ladd self-requested a review June 5, 2026 14:10

# How many consecutive cookieless probes with no new cookie before we stop.
# 10 is sufficient for Standard (2 pods) and Premium (4 pods); raise for larger fleets.
POD_DISCOVERY_TRIES=10

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

Consider increasing this to significantly decrease the chance of missing a pod on premium instances.

@benjermanjoel

Copy link
Copy Markdown
Author

Closing per Security request

@benjermanjoel benjermanjoel deleted the cloud-dp-inventory-refresh branch June 8, 2026 12:41
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants