Replies: 2 comments 3 replies
|
I haven't seen this exact case, but a couple of things you can check from your side:
Your plan to look at the December usage is a good test. If a line shows up for the deleted repo after the cycle resets, that's hard to blame on reporting lag. For the ticket, attaching the CSV rows tends to get further than describing the pattern. |
|
The number is more explainable than it looks, and the cause is probably the restore window. "GB-hr" is storage averaged over hours, not traffic. Your 42.04 GB-hr per day works out to roughly 1.75 GB being counted as present at every instant of the day — exactly the pace you're seeing during the day. So something about 1.75 GB is still considered stored. For a deleted repository, that's expected while it sits in the 90-day restore window. Deleting a repo doesn't immediately drop its data from the metering side; the snapshot is retained so it can be restored, and Actions storage is part of that. The exact repeated value every day is a good sign it's a fixed retained object rather than a runaway process. Two things worth doing:
You've already done the hard part by noticing the value is exactly constant. Keep that in the ticket — it rules out most billing bugs. |
Uh oh!
There was an error while loading. Please reload this page.
🏷️ Discussion Type
Question
💬 Feature/Topic Area
Other
Discussion Details
This is a follow-up to my earlier discussion: https://github.com/orgs/community/discussions/212674
The situation has developed since then, and the pattern is now much clearer, so I am opening a new discussion to ask if anyone else has seen this.
Background
What the usage page shows
GitHub's docs describe storage as metered hourly, so about 42 GB-hr per day means roughly 1.75 GB is still being counted as stored — for a repository that no longer exists.
What I have tried
Questions
I will report back after December 1 either way: if the metering stops at the cycle boundary, that supports a reporting/snapshot explanation; if a December line appears for a deleted repository, a cycle lag cannot explain it.
All reactions