Fix time zone normalization bug in interval schedules #19301
Merged
+59
−38
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
We discovered a regression in interval schedules after the pendulum → datetime/zoneinfo refactor. When the schedule JSON contains a fixed-offset anchor such as "2024-07-01T01:15:00+09:00" alongside a non-UTC timezone like "Asia/Tokyo", the scheduler drifts runs by the offset—so 01:15 JST became 10:15 JST. The culprit is the new tzinfo "normalization." For tzinfos that aren't ZoneInfo, we called `.tzname()` to get a timezone name; offset tzinfos return strings like "UTC+09:00", so we tried to create a ZoneInfo from that invalid key, which fails. The fix keeps offset tzinfos anchored to their original instant by projecting through dt.timestamp() into the target zone. Naive datetimes still assume the schedule's timezone, and true ZoneInfo objects continue down the existing path. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
CodSpeed Performance ReportMerging #19301 will not alter performanceComparing Summary
|
zzstoatzz
force-pushed
the
fix-interval-schedule-offset-anchor-timezone
branch
from
e656b91 to
87c28f1
Compare
desertaxle
approved these changes
Oct 29, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
We discovered a regression in interval schedules after the pendulum →
datetime/zoneinforefactor. When the schedule JSON contains a fixed-offset anchor such as"2024-07-01T01:15:00+09:00"alongside a non-UTC timezone like"Asia/Tokyo", the scheduler drifts runs by the offset—so 01:15 JST became 10:15 JST.Changes
The culprit is the new tzinfo "normalization." For tzinfos that aren't
ZoneInfo, the code called.tzname()to get a timezone name; offset tzinfos return strings like"UTC+09:00", which fails when trying to create aZoneInfofrom that invalid key.The fix uses
.astimezone(ZoneInfo(target_timezone))for offset-based tzinfos, which preserves both the absolute instant and microsecond precision. Naive datetimes still assume the schedule's timezone, and trueZoneInfoobjects continue down the existing.to_tz()path.Testing
Added regression test
test_offset_anchor_preserves_local_hourwhich:Test confirmed:
All existing interval schedule tests and scheduler service tests pass on Python 3.13.
🤖 Generated with Claude Code