My attempts to run `convex dev` fail to deploy code changes
When I run
npx convex dev -v --tail-logs, I get output about skipped files (the usual suspects about multiple dots and stuff in convex/_generated), and a raft of Preparing convex/** messages, but that's it β no messages about deployment.
Over in the dashboard, no updates in my Development deployment instance.7 Replies
Thanks for posting in <#1088161997662724167>.
Reminder: If you have a Convex Pro account, use the Convex Dashboard to file support tickets.
- Provide context: What are you trying to achieve, what is the end-user interaction, what are you seeing? (full error message, command output, etc.)
- Use search.convex.dev to search Docs, Stack, and Discord all at once.
- Additionally, you can post your questions in the Convex Community's <#1228095053885476985> channel to receive a response from AI.
- Avoid tagging staff unless specifically instructed.
Thank you!
And when I run the command with the
--once flag, I get a 1 exit status -- but no indication of any specific failure cause@prometheas π₯ what's your Convex version, and what's the backend you're deploying to? Or instead of answering the second question, could you verify your Discord / Convex handle
@prometheas π₯ it seems to be failing on the client, could you try
npx convex network-testHey @ballingt - had gone to my neighbor's comedy show last night, and then wasn't in "coding" state by the time the show was done π
I had first noticed this on 1.16.5 and tried upgrading to 1.17.0 seeing the same.
Additionally, today I'm finding a small change in the outcome β it gets to the bundling step when it fails:
Cursor suggested I look out for potential circular imports, but that seems like a red herring, since the
madge cli shows that there are circular dependencies between just about all of my convex/*.ts files and those that Convex generates in convex/_generated/**It sounds like the CLI is dying for some reason. Are you seeing this on e.g. a demo project instead of this project? What operating system are you on? Would you be able to share this code with us?
it's frustrating that no error is being printed, if the CLI dies in an unexpected way we'd expect to see a stack trace so something's wrong here.
Alright. So. Whatever tf is happening locally for me is NOT happening in a GitHub codespace that I just spun up, which is fun because in both cases I'm using a dev containerβ¦ so much for eliminating "works on MY machine" scenarios π€£ I'll keep digging and report back whenever I've gotten to the bottom of this
OK, I'm now back on track. I rebuilt my container (without cache), deleted
node_modules and ran npm i - something in that mix got me back in the game. A stack trace would ineed have been nice; I wonder what might have been going wrong to result in its presumably having been getting swallowed.
Anyhow - thanks again for your support, @ballingt!That's great! I'll take a look at cases that wouldn't print a stack trace for next time.