Hi!
I’ve noticed that the speed of Spin up environment (Docker)
step in our workflow varies 10x times over time.
The culprit is disk read/write speed.
Here are log messages from a normal run:
Status: Downloaded newer image for gcr.io//:***
pull stats: download 808.3MiB in 6.47s (124.9MiB/s), extract 925.3MiB in 15.826s (58.47MiB/s)
and 10x slower run:
Status: Downloaded newer image for gcr.io//:***
pull stats: download 925.6MiB in 12.211s (75.79MiB/s), extract 925.3MiB in 3m13.929s (4.772MiB/s)
Why does extract speed degrade 10x times ?
Is it a sign of hitting plan limits ?
Thanks in advance!