sneak/netwatch
1
0 Fork 0
Code Issues 5 Pull Requests 1 Actions 8 Packages Projects Releases Wiki Activity

The health box and summary wait for the round's slowest check, and say OFFLINE after the network is back #206

New Issue
Closed
opened 2026-11-11 21:16:30 +01:00 by clawbot · 1 comment
clawbot commented 2026-11-11 21:16:30 +01:00
Collaborator
Copy link
Copy source

Each row shows its result as soon as its check ends. The health box and the summary line update only once per round, after every check in it has ended (tick in web/src/main.js). So one target that never answers holds them back until the check timeout, 80% of the interval. For most users that happens every round:

  • Local CPE never answers for anyone without a cable modem at http://192.168.100.1, nor on a page served from a public address.
  • The Bahrain target never answers for anyone (#205).

The health box and summary then lag the rows by 2.4 s at the default interval, and by up to 48 s at the 60 s interval.

Found by the acceptance test on #204.

Steps: open the page and wait for HEALTHY. Take the browser offline until the box says OFFLINE, then bring it back.

What happens: within half a second of the network's return, the rows show green figures again. The health box still says OFFLINE, and Reachable says 0/26, for 2.5 s more, until the failing checks time out. Tested at the 3 s interval. The 48 s figure for the 60 s interval follows from the timeout rule and was not measured.

What should happen: the health box and summary never contradict the rows. They take in each WAN host's result when its row shows it.

Definition of done

  • The health box and summary update when a WAN host's check ends, not only when the whole round ends.
  • Checks still counts finished rounds.
  • A unit test: in a round where one WAN host never answers, the health box and summary reflect the hosts that did answer before that check times out.
  • README.md, Health and summary, still matches.

Model: opus-5-5

Each row shows its result as soon as its check ends. The health box and the summary line update only once per round, after every check in it has ended (`tick` in `web/src/main.js`). So one target that never answers holds them back until the check timeout, 80% of the interval. For most users that happens every round: - `Local CPE` never answers for anyone without a cable modem at `http://192.168.100.1`, nor on a page served from a public address. - The Bahrain target never answers for anyone (https://git.eeqj.de/sneak/netwatch/issues/205). The health box and summary then lag the rows by 2.4 s at the default interval, and by up to 48 s at the 60 s interval. Found by the acceptance test on https://git.eeqj.de/sneak/netwatch/issues/204. **Steps**: open the page and wait for `HEALTHY`. Take the browser offline until the box says `OFFLINE`, then bring it back. **What happens**: within half a second of the network's return, the rows show green figures again. The health box still says `OFFLINE`, and `Reachable` says `0/26`, for 2.5 s more, until the failing checks time out. Tested at the 3 s interval. The 48 s figure for the 60 s interval follows from the timeout rule and was not measured. **What should happen**: the health box and summary never contradict the rows. They take in each WAN host's result when its row shows it. ## Definition of done - [ ] The health box and summary update when a WAN host's check ends, not only when the whole round ends. - [ ] `Checks` still counts finished rounds. - [ ] A unit test: in a round where one WAN host never answers, the health box and summary reflect the hosts that did answer before that check times out. - [ ] `README.md`, Health and summary, still matches. Model: opus-5-5
clawbot commented 2026-11-11 21:29:39 +01:00
Author
Collaborator
Copy link
Copy source

Pull request: #217 (branch issue-206-health-box-per-check). The health box and summary now redraw as each WAN host's check ends. Checks still changes only at the end of a round. The Health: debug log line is still written once per round.

Model: opus-5-5

Pull request: https://git.eeqj.de/sneak/netwatch/pulls/217 (branch `issue-206-health-box-per-check`). The health box and summary now redraw as each WAN host's check ends. `Checks` still changes only at the end of a round. The `Health:` debug log line is still written once per round. Model: opus-5-5
Sign in to join this conversation.
No Branch/Tag Specified
Labels
Clear labels
critical
Extremely important: data loss, a security hole, production down, or basic use impossible
 merge-ready
needs-checks
needs-rebase
needs-review
needs-rework
No labels
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
clawbot sneak (Jeffrey Paul)
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: sneak/netwatch#206