Add --initial-retries for HEALTHCHECK
#26673
Conversation
This fix tries to address the issue raised in 26664 where it was not possible to have a grace period at the startup of the container. This fix adds `--initial-retries` for `HEALTHCHECK` so that container startup and normal running might have different grace period (retries). This fix adds an integration test to cover the changes. This fix fixes 26664. Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
|
We discussed this in the maintainers meeting, and while we think the health checks may need improvement, we are not sure about the right approach (yet). We prefer to wait one release to see what use cases people have, and issues they run into. For that reason I close this for now |
|
Has the option One solution could be to define 2 differents kind of healthchecks :
|
|
@sylvainmouquet docker 17.05 aded a During this period, health checks still run (so you can monitor them), but they are not "counted" as failures, hence, the service/container is not marked "unhealthy", and not restarted/tasks are not re-created if health check fails. See the pull request that implemented it #28938, and the associated issues (#26498 and #26664) for more information. |
- What I did
This fix tries to address the issue raised in #26664 where it was not possible to have a grace period at the startup of the container.
- How I did it
This fix adds
--initial-retriesforHEALTHCHECKso that container startup and normal running might have different grace period (retries).- How to verify it
This fix adds an integration test to cover the changes.
- Description for the changelog
Add
--initial-retriesforHEALTHCHECK, so that it is possible to specify an initial grace period that is different from the normal grace period (--retries).- A picture of a cute animal (not mandatory but encouraged)
This fix fixes #26664.
Signed-off-by: Yong Tang yong.tang.github@outlook.com