GraphQL API Issue: Launching Pod Fails (GRAPHQL_VALIDATION_FAILED)
I'm unable to launch a specific pod configuration via the GraphQL API (podFindAndDeployOnDemand at https://api.runpod.io/graphql), consistently getting an HTTP 400 GRAPHQL_VALIDATION_FAILED error.
My myself query works fine with the same API key.
Target Configuration:
GPU: NVIDIA A40
Image: runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04
Network Volume ID: hlm2fs1bpr (at /workspace)
Ports: 8888/http, 8188/http, 22/tcp
Container Disk: 20 GB
Attempts:
Using templateId="runpod-torch-v240" (like the UI sends) -> fails validation.
Using imageName with various input parameters (minimal, mimicking UI details like minMemoryInGb, dataCenterId, etc.) -> also fails validation.
Example Failed Input (using imageName):
{
"input": {
"imageName": "runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04",
"gpuTypeId": "NVIDIA A40",
"containerDiskInGb": 20,
"ports": "8888/http,8188/http,22/tcp",
"networkVolumeId": "hlm2fs1bpr",
"volumeMountPath": "/workspace"
}
}
Question:
What is the correct input structure required by the API to launch this specific configuration successfully?
12 Replies
might needto use templateId instead
or try the rest api
https://rest.runpod.io/v1/docs#tag/pods/POST/pods
its newer, i bet its working better
@Jason Tried it, nothing works from this, I couldn't find the id but it seems to be: runpod-torch-v240
I found it through inspect
i see
the rest api also doesnt work?
nah(
what does it return? i wanna take a look at the response
@Jason what shows me terminal
2025-04-13 17:04:10,337 - main - INFO - Starting bot...
2025-04-13 17:04:10,645 - telegram.ext.Application - INFO - Application started
/Users/user/user/runpod_telegram_bot/.venv/lib/python3.13/site-packages/telegram/ext/_handlers/conversationhandler.py:858: PTBUserWarning: No 123456.
2025-04-13 17:04:20,146 - main - INFO - User 123456 initiated MINIMAL pod launch with image runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04, GPU NVIDIA A40
2025-04-13 17:04:20,581 - main - ERROR - HTTP Error 400: {"errors":[{"message":"Something went wrong. Please try again later or contact support.","locations":[{"line":1,"column":28}],"extensions":{"code":"GRAPHQL_VALIDATION_FAILED"}}]}
2025-04-13 17:04:20,582 - main - ERROR - Pod launch failed: Ошибка HTTP 400, Details: Something went wrong. Please try again later or contact support.
^C2025-04-13 17:04:34,766 - telegram.ext.Application - INFO - Application is stopping. This might take a moment.
and it happens no matter what I do
JobQueue set up. To use JobQueue, you must install PTB via pip install "python-telegram-bot[job-queue]".
if application.job_queue is None:
/Users/user/user/runpod_telegram_bot/.venv/lib/python3.13/site-packages/telegram/ext/_handlers/conversationhandler.py:859: PTBUserWarning: Ignoring conversation_timeout because the Application has no JobQueue.
warn(
2025-04-13 17:04:18,692 - main - INFO - API Key stored for oh that is using the rest api or graphql??
in this case is graphql, but with rest api same
ah so, the error from rest api doesnt say anything helpful?
yeah @Jason
i think you might wanna try in curl or python / whatever language to test the request for creating pods
no need to ping me everytime hahah
chill
sorry haha
okay, thanks, I'll try again