-
Notifications
You must be signed in to change notification settings - Fork 518
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Unable to build server #2841
Comments
Can you please share a runnable reproduction? 🙏 |
I may be able to contribute an MRE, I encountered a similar issue in Solid Start. However, the build still completes, just with a ton of duplicated messages identical to the one above, and takes about 2x as long to build (before this, build times were pretty consistent). |
Turns out the Solid Start bare starter reproduces this, fun! Made a quick MRE: https://github.com/agoldstein03/nitro-issue-2841-mre Full logs
|
I added a Full logs
|
Thanks for reproduction dear @agoldstein03! #2842 should fix it. |
Thanks for the quick PR @pi0! Also curious, is this something that a test could catch in the future before it's published? I'm unfortunately not familiar enough with the codebase, sorry. I could see how these types of warnings would be less stable to test for, but it seems like a regression test for build performance might be worth it to catch if build times suddenly double. Thoughts? |
To be honest i believe it is something could be caught using rollup types. If passing argument is necessary, it should have been required. But also we don't have many usecases of this so test is probably not necessary. |
@pi0 My original message was not very clear, apologies. I understand this specific rollup issue is unlikely to occur again, but since it also had a significant negative impact on build times (which is also undesirable), it seems to me that adding general tests to track whether build times stay within some generous window (+/- 25% expected, for example, to account for runtime variation) even just before releases might be worth it. This would provide early warning for any future issues which the tests don't otherwise catch but also happen to negatively impact build times as a byproduct (and trigger a closer look at logs, for example), in addition to testing performance regression. As best as I can tell, the current tests do include a production build, but only test whether or not they build successfully or error. Would an additional check like as described above for build duration be considered? |
It might be a viable thing to do once we have an ecosystem CI. diff in this case was so low (1.5s => 1.6s on my system) and most of the time it will be noticed when using Nitro in more complex setups only. |
Environment
Operating System: Linux
Node Version: v20.15.1
Nuxt Version: 3.13.2
CLI Version: 3.15.0
Nitro Version: 2.10.1
Package Manager: bun@1.1.34
Builder: -
User Config: default
Reproduction
Describe the bug
When I try to build my Nuxt project I get the next warning and then the build crashes:
I was able to build my project before, I think is related with commit 51faaf3
Additional context
No response
Logs
No response
The text was updated successfully, but these errors were encountered: