How Dockerfile Layers/Caching Work

This video provides a quick overview of how the different commands in your Dockerfile translate into layers in the image. We take a nieve Dockerfile and re-write it to take advantage of caching to get faster builds and a better development experience.

[snizamaddinov]

Great video. Thank you for a straightforward explanation.

[peterbergman]

I like how you included a practical example of where this knowledge often comes in handy. Thank you, sir.

[skies]

I just want to pinpoint that tools like JIB, always put your code as the last layer, so that it builds only that. That is : for a well structured java app (the vast majority), they will have the layers build exactly like you showed, for free, without users even knowing that. This was a significant improvement for us.

Also tools like kpack, try to mitigate the order at all. So for them, order does not even matter, which I find "magical". thank you for the short and good video.

[bcr]

Thanks broo! Very good content. Watching from Brazil!

[nareshg]

Simple and easy explanation thanks man

[nikosmarshmallow]

Excellent video and very well explained.

[strandingstranger]

Thank you so much! Loved the explanation.

[tryptamigo]

fun fact: git actually isn’t built on diffs but rather complete snapshots. every commit ref is a full copy of the code base.

[chewbucca]

Thanks mate that was a good explanation

[bill_carroll]

Just a friendly reminder that git doesn’t store diffs 😊

[rhll]

Still not able to comprehend that why it doesn't use the cache when the same set of commands are changed in the order in which they occur, for e.g. the set of commands that installed dependencies where when shifted above the WORKDIR started to use cache which was not the case earlier.

[MegaOsama]

wow amazing thanx too much bro you helped me

[xtazyxxx]

What if you using a microsoft hosted agent in azure devops pipeline building the docker file. I assume as they are different vm everytime thus caching would not work