Erik Sipsma

My first guess (truly just a hunch though) at the error hit in @aluzzardi's example is that the system is running out of memory and processes are getting OOM killed....

> When I tried to reproduce @sipsma, I didn't get any OOM's, just a very slow start, and a slow progress. Most of it happened **after** `dagger` CLI exited. Happy...

Okay yeah the 5 minute hang on OTEL draining repros consistently on main for me: ``` ./hack/dev go test -v -count=1 -timeout=1h -run=TestFileServiceSync/triggers_error ./core/integration/ ``` Funnily if you `SIGQUIT` or...

@gerhard the SIGQUIT stack you attached in the description works in swirly for me, the logs seem to have one stack trace, then logs from the engine running again and...

@gerhard summarized a lot of the tradeoffs I would have mentioned 🙂 So I'll just add some more thoughts --- For users already invested in a horizontal scaling architecture, they...

> as long as the local cache fits on an EBS volume which can be the only resource you persist). @samalba This may seem like a small detail, but it's...

Fixed by https://github.com/dagger/dagger/pull/10435

Yeah I brought up something similar here in a discussion about making exports more modular: https://github.com/moby/buildkit/issues/3037#issuecomment-1235034356 I like this general idea a lot. Between all the code in `cache` for...

I'm curious to know the specifics of the use case here, especially around how the data in the NFS mount would be used. The two categories of data we support...

> having the ability to mount FFS shares in dagger would solve our problem but I don't think this is practical to implement. The environment in which we run won't...