Timeout Issue Despite Response Within Timeout Limi...
# support
s
Hi Team, I'm facing an issue where vapi.start(options, { silenceTimeoutSeconds: 120 }) is used to extend the timeout duration. The tool-call webhook is triggered correctly and receives all inputs, but for some functions that take a bit longer (yet still respond within 120 seconds), I encounter the following error:
[ERROR] Your server rejected 'tool-calls' webhook. Error: Request failed with status code 504.
It seems like VAPI isn't waiting for the tool-call response as expected, even though everything is within the timeout limit. Also, please confirm if I'm using the timeout parameters correctly. Here's how I’m currently setting them:
vapi.start(options, {silenceTimeoutSeconds: 120});
And for the tools config:
tools: {server: {url: "server_url",timeoutSeconds: 120}}
Please let me know if this is the correct way to configure tool-call timeouts or if there’s anything I’m missing. You can refer the call id
650c36fd-796f-4ed4-86d4-da8e4c0dca77
which describs this error. Thanks, Sahil
@Vapi team would really appreciate a quick support on this
s
Looking at the log, I can confirm that you've correctly configured the timeout settings:
Copy code
json
"server": {
  "url": "https://axlai.accelnomics.com/api/configuration/webCall/67bdb5d69515b737a53ee869",
  "timeoutSeconds": 120
}
Your tool is correctly configured as synchronous (
"async": false
) with a 120-second timeout, which should allow enough time for longer-running operations. ## What's Likely Happening Despite your VAPI configuration being correct with the 120-second timeout, you're still receiving a 504 Gateway Timeout error. This strongly suggests that the timeout is occurring at an infrastructure level, outside of VAPI's control. There are several possibilities: 1. **Intermediate Proxy/Gateway Timeout**: Your API might be behind a proxy, API gateway, or load balancer that has its own (shorter) timeout setting, likely around 60 seconds. 2. **Cloud Platform Timeout**: If your webhook is hosted on a serverless platform like AWS Lambda, Google Cloud Functions, or Azure Functions, these platforms often have default timeout settings that may need to be increased. 3. **Web Server Timeout**: The web server hosting your webhook might have a default timeout that's lower than 120 seconds.
@Sahil For fair consideration, I will request you to share another call ID when I'm asking for another call ID. This will help me to create a pattern or reference where this is happening for more than one call, and then I can ask the team to investigate on this.
s
Hi Shubham appreciate your quick support on this. Reading your observations it seems there's some issue in our web server or Proxy/Gateway I will double check and comeback
v
@sahil_17638 following up on this, if you had a chance to look into your logs or did you find something on your side?
s
Hi Shubham Thanks for following up. It seems this is the issue on our side only that we need to fix
v
Hey Sahil, Got it, I'll close out this ticket now then. Let me know if you guys need any support or have questions from your side of things!
s
ok
5 Views