0

I know when dynamo throttles requests and that the AWS sdk retries a throttled request until it succeeds and that apparently a error is returned when dynamo cannot handle more requests. I also read about all the possible dynamoDB errors. Based on this, I deduce that a throttled request is mostly causing latency but no errors/ bad request. Is this correct? How can I see dynamoDB errors due to too much throttling?

Could that kind of error be seen in the error GUI of the dynamoDB table (see image)? enter image description here

DaveR
  • 1,696
  • 18
  • 24

1 Answers1

0

These are related to an error returned from the DynamoDB API, for which a 500 error is returned.

The requests to DynamoDB or Amazon DynamoDB Streams that generate an HTTP 500 status code during the specified time period. An HTTP 500 usually indicates an internal service error.

Whilst this could include throttling, it could also be used in an event where the DynamoDB service is not functioning correctly within the region. It will ultimately be an issue with the service rather than your requests, although your applications behaviour will be the cause in a throttling situation.

More information is available in the DynamoDB Metrics and Dimensions documentation.

Chris Williams
  • 32,215
  • 4
  • 30
  • 68
  • If I understand correctly that link and what you said, it means there is no way exactly to see how many throttled requests receive an error right? One can only 'deduce' that those errors are generated by throttled requests – DaveR Feb 10 '21 at 11:42
  • There is a separate metric for specifically just throttles: "ThrottledRequests" – Chris Williams Feb 10 '21 at 11:45
  • Yes I know about that metric, my question how/where can I see if those "ThrottledRequests" are actually generating errors? – DaveR Feb 10 '21 at 13:54