Seeking advice on unexpected Dagger cache growth (~50GB per run)

[frail creek]

Hi everyone! I’m investigating a storage issue on our CI server (GitLab runner) and would appreciate some guidance.

Even though the project and its dependencies are only a few GBs, our Dagger cache mount grows by roughly 50GB during a single CI run. I'm trying to determine if this is expected behavior or an optimization issue on our end.

Specifically, I’m looking for:

• Tools or commands to inspect which layers/outputs are consuming the most space in the cache.
• Best practices for Dagger functions to avoid unnecessary layer duplication.
• Any common "gotchas" that might cause outsized cache growth.

Thanks in advance for any insights!

[edgy sapphire]

hey tony! have you checked the dagger core engine local-cache command? that has some very useuful information about the data entries in the cache along with their metadata

[frail creek]

I have looked at dagger core engine local-cache entry-set where I can see the size but I am not sure how to use that information. I can see the size but I don't know "why" it is growing so fast

I will look into that command some more, thanks

[edgy sapphire]

we've added the record-type in the latest version of the engine which tells what type of cache entry record that is. That along with the description field should give a good idea about where that cache entry is coming from