[GH-ISSUE #910] Alert if /start signal does not arrive at expected time (with a separate grace time setting for it) #637

New issue

Open

opened 2026-02-25 23:43:07 +03:00 by kerem · 4 comments

kerem commented

2026-02-25 23:43:07 +03:00

Owner

Originally created by @haikosaw on GitHub (Oct 28, 2023).
Original GitHub issue: https://github.com/healthchecks/healthchecks/issues/910

I looked around a bit but it didn´t seem like anyone spoke about this sort of thing yet.

Currently if I want to use the /start ping, the grace period dictates by what time a finish ping has to arrive. My use case would benefit from a distinction between a grace period for "job completion", and "job start".

I´m not sure what the best way would be of implementing such a thing, but what it would need to do is:

Send an alert if the task is not run within the start schedule (example 1 minute after schedule)
But also send alert when task is exceeding it's runtime (example grace configured for 1 hour, but no finish ping has come yet)

Originally created by @haikosaw on GitHub (Oct 28, 2023). Original GitHub issue: https://github.com/healthchecks/healthchecks/issues/910 I looked around a bit but it didn´t seem like anyone spoke about this sort of thing yet. Currently if I want to use the /start ping, the grace period dictates by what time a finish ping has to arrive. My use case would benefit from a distinction between a grace period for "job completion", and "job start". I´m not sure what the best way would be of implementing such a thing, but what it would need to do is: - Send an alert if the task is not run within the start schedule (example 1 minute after schedule) - But also send alert when task is exceeding it's runtime (example grace configured for 1 hour, but no finish ping has come yet)

kerem added the

feature

label

2026-02-25 23:43:07 +03:00

kerem commented

2026-02-25 23:43:08 +03:00

Author

Owner

@moritzruth commented on GitHub (Mar 20, 2024):

My use case for this would be a backup script that

runs every day at the same time (start grace time: 1min)
takes multiple hours to run (completion grace time: 6 hours)

@moritzruth commented on GitHub (Mar 20, 2024): My use case for this would be a backup script that - runs every day at the same time (start grace time: 1min) - takes multiple hours to run (completion grace time: 6 hours)

kerem commented

2026-02-25 23:43:08 +03:00

Author

Owner

@malarzm commented on GitHub (Jun 27, 2024):

It'd be also great to allow "no limit" for a cronjob run. In our case we're interested in having a run id (for log correlation) and time the cron actually ran.

@malarzm commented on GitHub (Jun 27, 2024): It'd be also great to allow "no limit" for a cronjob run. In our case we're interested in having a run id (for log correlation) and time the cron actually ran.

kerem commented

2026-02-25 23:43:08 +03:00

Author

Owner

@amcsi commented on GitHub (Oct 30, 2025):

I think #547 may be related. It would be nice to separately watch start times.

@amcsi commented on GitHub (Oct 30, 2025): I think #547 may be related. It would be nice to separately watch start times.

kerem commented

2026-02-25 23:43:08 +03:00

Author

Owner

@Vyerni commented on GitHub (Nov 14, 2025):

It'd be very helpful to have a grace period of a "late start" different to a grace period of a start signal received

@Vyerni commented on GitHub (Nov 14, 2025): It'd be very helpful to have a grace period of a "late start" different to a grace period of a start signal received