Create a diagnostic to warn if the durations in HEALTHCHECK's flags are too short

[rcjsuen]

Docker doesn't allow the times in the interval, start-period, or timeout flags of HEALTHCHECK instructions to be less than one millisecond. We should create a diagnostic to check for this case.

FROM scratch
HEALTHCHECK --timeout=100ns CMD ls

$ docker build .
Sending build context to Docker daemon  86.15MB
Step 1/2 : FROM scratch
 --->
Step 2/2 : HEALTHCHECK --timeout=100ns CMD ls
Interval "timeout" cannot be less than 1ms

[rcjsuen]

Valid units of time are: "ns", "us" (or "µs"), "ms", "s", "m", and "h".

As aforementioned, the total sum must be greater than or equal to one millisecond.

FROM alpine
HEALTHCHECK --interval=1ms --start-period=1ms --timeout=1ms CMD ls

$ docker build .
Sending build context to Docker daemon  86.15MB
Step 1/2 : FROM alpine
latest: Pulling from library/alpine
88286f41530e: Pull complete
Digest: sha256:1072e499f3f655a032e88542330cf75b02e7bdf673278f701d7ba61629ee3ebe
Status: Downloaded newer image for alpine:latest
 ---> 7328f6f8b418
Step 2/2 : HEALTHCHECK --interval=1ms --start-period=1ms --timeout=1ms CMD ls
 ---> Running in b380b974e4de
 ---> 4c532672bbd0
Removing intermediate container b380b974e4de
Successfully built 4c532672bbd0