Skip to content

Conversation

@Mark-Simulacrum
Copy link
Member

@Mark-Simulacrum
Copy link
Member Author

Review nits fixed and verified that this works locally (by setting allowed disk usage to 20% or so)

@onur
Copy link
Contributor

onur commented Jul 30, 2019

thanks @Mark-Simulacrum but I think 20% might not be enough. Some crates (stm32) generate more than 10 GB documentation (no idea about their build artifacts) and since our disk space is really limited now, maybe we can increase it to at least 30 or 40 percent.

This avoids filling up the disk and leading to postgres failure.
@Mark-Simulacrum
Copy link
Member Author

We're currently sitting at 50 GB free on the server (out of 100 GB total). The way the logic is implemented, all 50 GB can be used by a single crate compilation (AFAICT), but before/after each crate build we will clear out the target directory if the usage exceeds 80GB. So, basically, this isn't making things any worse for an individual crate compilation.

@onur
Copy link
Contributor

onur commented Jul 30, 2019

Then its great 👍

@onur
Copy link
Contributor

onur commented Jul 30, 2019

I am merging this before we ran out of space again 😂

@onur onur merged commit a313440 into rust-lang:master Jul 30, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants