Deployments keep failing (build process exit: missing Sentry binary)

Help
,
1

3 days ago I did a deployment that succeeded, but two days ago the first one failed and since then I have not been able to deploy one successfully, if I do a redeploy of a build that previously succeeded then it still works but redeploying the ones that failed still fails

The error I usually get is:

unhandledRejection Error: Sentry CLI binary for this platform/architecture not found!
at new Promise ()
Error: Command “npm run build” exited with 1

What bugs me, is that during the npm installation process there are no errors in the logs, the missing binary error pops up later when Next.js has started the build process, as I can see the line that reads: “Creating an optimized production build …”

So I checked the Sentry Javascript SDK issues, the Sentry CLI issues, the Vercel and Next.js issues, forums and discussion boards, somehow it seems like I’m the only one impacted :confused:

I noticed that the last build that worked had Next.js 15.0.4-canary.52 (the ones that failed has 15.1.1 and some of the canaries that followed), so I did a new deployment using that version but it failed too

My code is public, here is the commit list for the review branch, which also shows you for each commit if the deployment was successful or not: Commits · chrisweb/chris.lu · GitHub (I have looked at my commit history, especially the ones around the time the first builds started failing but I can’t find a problem)

Between the date where builds still succeeded and the date the first one failed, I changed the nodejs version via the vercel interface to 22 (previously 18)

My builds on local (windows 10) are successful (but then the same code produces a build error on vercel)

So yeah I’m running out of ideas, does anyone have an idea what could be wrong?

2

Please check if the sentry-cli binary is located in @sentry/cli/bin.

In any case. Vercel caches builds and dependecies. Can you try rerunning a build without cache? There’s a checkbox in Vercel for that.

1 Like
3

I was attempting to answer your question and did a new deployment, now I get completly different errors:

npm error code EINTEGRITY
11:08:30.288
npm error sha512-34V5v2qmeBdl34wlV0dMJMJFc5mqQvNBokti/7DLOC/qYflRdW12uhvmwQCJdamR+ccW0BJC+pMo7/avcLcDAQ== integrity checksum failed when using sha512: wanted sha512-34V5v2qmeBdl34wlV0dMJMJFc5mqQvNBokti/7DLOC/qYflRdW12uhvmwQCJdamR+ccW0BJC+pMo7/avcLcDAQ== but got sha512-1odDWP7MRRpP+cc79DkLnct7+mia1WBvRSJ8e+XFF/Yhy97EA9qEza2XqckcvvIJY++owUf9kCAuJ5moABOQ3Q==. (4809 bytes)
11:08:30.289
npm error A complete log of this run can be found in: /vercel/.npm/_logs/2024-12-16T10_06_03_079Z-debug-0.log
11:08:30.325
Error: Command "npm install" exited with 1

EDIT 1: By run a new build, I assume you mean do a “redeploy” and leave the “use cache from previous build” unchecked!?

EDIT 2: Hmmm I don’t know how I can see the build source on vercel? When I go to the deployments list I have a “view source” option, but that seems to be my repository source and not the build process source

EDIT 3: did a redeploy after the first “npm error code EINTEGRITY” and get the same errors again, so I’m going to downgrade next.js by a minor and do another deployment

4

I wonder if your package-lock.json version is corrupted. Can you remove it and regenerate a fresh one to see if it helps?

5

Downgrading nextjs caused the same error about corrupt files

Hmmm that would explain the problems, will delete it and retry

6

Nice deployment successful :slight_smile: thank you @swarnava

last problem with hash integrety was due to me altering the lock, my bad :frowning: but the error I had the previous day regarding sentry, was that something you fixed on your side?

[done] regarding the checking the files of a build, can I for example check is in the node_modules folder, no matter if a build failed or was successful, or is only you guys who can see that? → the answer is here Inspect your deployment source and build output files - Vercel

1 Like
7

You can only view files for deployment that are successfully built.

In general, we don’t have access to deployment files unless customer enable it for debugging. :slight_smile:

Glad to heat this is resolved now!

8

ah only successful ones, I get it

only when allowed, makes sense, added the page to my vercel bookmarks, you never know

thank you for answering all my remaining questions :slight_smile:

2 Likes
9

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.