Hasura
Hi,
I am getting this error while trying to get Hasura running:
Error checking if container is ready: connect(): Connection refused: container port not found. Make sure you exposed the port in your container definition.
✘ [ERROR] Error checking if container is ready: The operation was aborted
Hasura container error: [Error: Container exited with unexpected exit code 139]
I have tried pushing to the registry as described in Image Management on the Docs.
Any help much appreciated!
12 Replies
Hey @various looks like the process in the container itself is crashing?
Anything interesting in the Containers Logs tab for that deployment?
Hey Thomas, thanks so much for replying so quickly.
What I pasted above was from local, after deploying, the Containers Logs read:
Error checking 8080: The container is not listening in the TCP address 10.0.0.1:8080.
and
"Error: A call to blockConcurrencyWhile() in a Durable Object waited for too long. The call was canceled and the Durable Object was reset.
Early today I was able to get it working locally, but then it didn't work when I deployed. I did some more reading of the docs and learnt that it has to be linux/amd64. I tried to change to that and that was where the problems started locally. Unfortunately I didn't commit anything (seemed so easy to get it working locally) and I don't know where the problem is now.
I tried again from scratch twice and couldn't get to where I was.
I also can't figure out how to delete whatever I already pushed to the registry to give myself a clean slate.
I have been experiencing the exact same issue since yesterday. I haven't heard back from CF team, though
https://discord.com/channels/595317990191398933/1387397692702982256/1387504214363734178
Hey @cdrpico, thanks for linking
I also can't figure out how to delete whatever I already pushed to the registry to give myself a clean slate.```You can use
npx wrangler containers images delete
Can you make sure the app in the container is listening on 0.0.0.0:8080 and not on localhost:8080 ?
@cdrpico I see uvicorn logs it is listening on 0.0.0.0:8000 so you should be goodYes, in my case it works but it takes up to 2 mins for the app to actually get ready to start receiving requests
I think that, because I don't know how to get the id of hasura that I pushed to the registry (it doesn't show when I run
npx wrangler containers images list), I can't. How do I get the id?
-- The only things I see are the images I have specified in a wrangler.json somewhere.In the docs for Hasura it says it listens to everything by default. Otherwise I am not sure what you mean.
(https://hasura.io/docs/2.0/deployment/graphql-engine-flags/reference/#server-port)
GraphQL Engine Server Configuration Reference | Hasura GraphQL Docs
Hasura GraphQL Engine Server Configuration Reference (Flags/Environment variables)
Hi @Thomas Lefebvre, should I make a github issue for deleting such images? Or, where do I go from here?
I think https://discord.com/channels/595317990191398933/1387061528016195684/1390095914680717446 #2 and #3 should help with hasura and uvicorn start up time and no OOM
Can you please link your Cloudflare account with your Discord user with /link so I can take a look why you are not seeing your images in list?
How were these images pushed by the way? Always via wrangler deploy?
I connected the dots with another bug report of images pushed with
npx wrangler containers push not showing up in list (only images pushed via deploy showing up) and this should be fix on https://github.com/cloudflare/workers-sdk/pull/9811
You can test yourself with npm i https://pkg.pr.new/wrangler@9811 and npx wrangler@https://pkg.pr.new/wrangler@9811 containers images listGitHub
wrangler containers images delete should prefix the account id by d...
Fixes CC-5572.
Fixes #9837
When listing container images with wrangler container images list, we omit the account ID prefix that each container image has in the managed registry. When deleting imag...
Hey Thomas, thank you so much for your work!
I installed the wrangler you linked.
When I run
containers images list, I can now see the remote hasura image in the list. However, it is possible I did not give you 100% accurate information before when I said it wasn't showing. Maybe I was just confused -- I was looking for some "ID" so as to delete the image, and maybe when I wrote the message I ran containers list expecting it to be there, as the cli clearly shows an ID for each container when you run this (but not for containers images list, and hasura is not listed there, so I thought okay I can't get an id if I can't see it.
But just now I made a guess as to what the ID might be (repo:tag) and it worked! So fantastic, I will give it another go now. Thank you