ProbeForever has 1 second delay

[jsafrane]

Originally reported by @Fricounet in #175.

The previous beheavior of ProbeForever was to try a probe and then sleep for the probeInterval duration. When it was modified to use a time.Ticker instead of a sleep in #149, it changed the behavior to wait the probeInterval first and then do the probe.

This adds unnecessary 1 second delay for the first Probe() call.

[jsafrane]

/close

This is fixed by #177. I'm using this issue to track what CSI sidecars need update.

[k8s-ci-robot]

@jsafrane: Closing this issue.

In response to this:

/close

This is fixed by #177. I'm using this issue to track what CSI sidecars need update.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[jsafrane]

Needs update + patch release, because they use ProbeForever:

• external-attacher
• external-health-monitor
• external-provisioner
• external-resizer
• external-snapshotter

Does not use ProbeForever:

• livenessprobe
• node-driver-registrar