Debugging highly variable skopeo copy times

[cweld510]

I'm seeing highly variable latencies when running skopeo copy to pull down a particular image from a remote registry, even when increasing parallelism. Does skopeo expose any internal stats or measurements that I could use to debug this problem, and help determine where the bottleneck is?

[mtrmac]

Thanks for reaching out.

There is --debug; that can be used to discover avoided copies (the principal expected source of variability) and retries (with unreliable servers/links), and maybe to time things, although it’s probably not very convenient for that.

[github-actions]

A friendly reminder that this issue had no activity for 30 days.

[mtrmac]

Please reopen if there is are more outstanding questions.