cluster wide property aws.requestTimeoutMillis set to 900,000 milliseconds (15 minutes). Despite this, we are still seeing the AWS assertion throw the error:
{"Error":"com.amazonaws.SdkClientException: Unable to execute HTTP request: Read timed out"}
when the lambda runs for more than 200,000 ms.
We are running AWSSolutionKit version 1.0.01
On the AWS side, the lambda is configured with maximum runtime of 15 minutes, and CloudWatch logs show it successfully completes execution each time,
so there's no timeout happening on the AWS side.
Release : 10.0
Component : API GATEWAY
1. These are the 4 new CWP we have introduced
aws.requestTimeoutMillis
aws.socketTimeoutMillis
aws.connectionTimeoutMillis
aws.executionTimeoutMillis
2. for the current Issue aws.socketTimeoutMillis is the fix, by default this will be 50Seconds provided by AWS
3. Customer have now the capability to configure it using the CWP.
The timeout was happening because of this socket timeout and now by increasing it they will not run in to this issue.
4. Provided new Fix : AWSSolutionKit-1.0.02.sskar
5. Raising aws.socketTimeoutMillis =900000 got rid of the error for delay=300000&bytes=250000
1. Fix is production ready , generated out of official teamcity builds and uploaded in our artifactory
2. Please open a Support case to obtain the fix while is added to the current version