Continuous watching by reusing the resourceVersion multiple times

[nolar]

Optimize the K8s-watching reconnections by valid resource version

Issue: #223

Description

Currently, Kopf does 1 listing and 1 watching query per iteration: the listing is to get the latest valid resource version, from which the watching continues. The listed items are simulated to be yielded as if from a watch-stream.

When the watch-query disconnects (this happens every ~60 seconds if the timeout is not set), the whole list is re-read again. This leads to @kopf.on.resume handlers being called multiple times for no actual need.

With this PR, the watch-queries will be repeated as long as the known latest resource version value is valid (i.e. accepted by the API). Once 410 GONE error is returned, the stream will stop and escalate to the full re-listing as before.

Types of Changes

• Bug fix (non-breaking change which fixes an issue)
• Refactor/improvements

Review

List of tasks the reviewer must do to review the PR

• Tests

[zincr]

🤖 zincr found 0 problems , 0 warnings

✅ Large Commits
✅ Approvals
✅ Specification
✅ Dependency Licensing

[zincr]

🤖 zincr found 1 problem , 0 warnings

❌ Approvals
✅ Large Commits
✅ Specification
✅ Dependency Licensing

Details on how to resolve are provided below

---

Approvals

All proposed changes must be reviewed by project maintainers before they can be merged

Not enough people have approved this pull request - please ensure that 1 additional user, who have not contributed to this pull request approve the changes.

• ✅ Approved by PR author @nolar
• ❌ 1 additional approval needed