aws-lambda: Log retention gives rate exceeded error

[Exter-dg]

Describe the bug

Legacy log retention in Lambda gives a rate limit exceeded error.

We are in the process of upgrading our app from CDK v1 to v2. To test this, we created a new env in a new account and redeployed the configuration using CDK v1.

We are creating 70-80 lambdas with log retention enabled. The legacy log retention creates a custom lambda resource to create log group and set log retention. CDK V1 used to create Node 14 lambas for this purpose (for which the creation is blocked in AWS). Hence, we disabled the log retention and upgraded the stack to 2.151.0 and then enabled the log retention.

While doing so, our stack is failing with the error:

Received response status [FAILED] from custom resource. Message returned: Out of attempts to change log group

Initially we thought this is an issue with the “CreateLogGroup throttle limit in transactions per second” quota. We increased it to 80 from 10 but the issue still exists.

On exploring the cloudwatch logs for the custom lambda resource, we found:

2024-09-06T05:23:33.260Z	06a9833f-0ad3-4faf-8f94-aa78dd49d0ec	ERROR	{
  clientName: 'CloudWatchLogsClient',
  commandName: 'PutRetentionPolicyCommand',
  input: {
    logGroupName: '/aws/lambda/LogRetentionaae0aa3c5b4d-mE6Tt6xks1CB',
    retentionInDays: 1
  },
  error: ThrottlingException: Rate exceeded
      at de_ThrottlingExceptionRes (/var/runtime/node_modules/@aws-sdk/client-cloudwatch-logs/dist-cjs/index.js:2321:21)
      at de_CommandError (/var/runtime/node_modules/@aws-sdk/client-cloudwatch-logs/dist-cjs/index.js:2167:19)
      at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
      at async /var/runtime/node_modules/@aws-sdk/node_modules/@smithy/middleware-serde/dist-cjs/index.js:35:20
      at async /var/runtime/node_modules/@aws-sdk/node_modules/@smithy/core/dist-cjs/index.js:165:18
      at async /var/runtime/node_modules/@aws-sdk/node_modules/@smithy/middleware-retry/dist-cjs/index.js:320:38
      at async /var/runtime/node_modules/@aws-sdk/middleware-logger/dist-cjs/index.js:34:22
      at async /var/task/index.js:1:1148
      at async /var/task/index.js:1:2728
      at async y (/var/task/index.js:1:1046) {
    '$fault': 'client',
    '$metadata': {
      httpStatusCode: 400,
      requestId: 'e247739c-8ebb-40d3-b85e-293802a87e24',
      extendedRequestId: undefined,
      cfId: undefined,
      attempts: 3,
      totalRetryDelay: 466
    },
    __type: 'ThrottlingException'
  },
  metadata: {
    httpStatusCode: 400,
    requestId: 'e247739c-8ebb-40d3-b85e-293802a87e24',
    extendedRequestId: undefined,
    cfId: undefined,
    attempts: 3,
    totalRetryDelay: 466
  }
}

Looks like an issue with the rate limit for PutRetentionPolicyCommand. The service quota for the same cannot be changed. Our earlier implementation had one difference in how log retention was implemented.
The base property was enabled to apply a exponential backoff (probably to handle such cases). This is now deprecated and hence we removed it during our upgrade from CDK v1 to v2. The documentation for LogRetentionRetryOptions says that this was removed as it is handled differently in AWS SDK v3. Is this what is causing the issue? Should't CDK/ SDK handle the backoff in this case?

Regression Issue

• Select this option if this issue appears to be a regression.

Last Known Working CDK Version

1.204.0

Expected Behavior

Log retention backoff should be handled internally

Current Behavior

Creating legacy log retention for multiple lambas together gives a rate limit exceeded error.

Reproduction Steps

Possible Solution

No response

Additional Information/Context

No response

CDK CLI Version

2.108.1

Framework Version

No response

Node.js Version

v22.1.0

OS

MacOs

Language

TypeScript

Language Version

No response

Other information

No response

[Exter-dg]

@rix0rrr Is this related?
#24485

[Exter-dg]

We fixed it by increasing the value of maxRetries from 7 to 20.

[github-actions]

Comments on closed issues and PRs are hard for our team to see.
If you need help, please open a new issue that references this one.

[github-actions]

Comments on closed issues and PRs are hard for our team to see.
If you need help, please open a new issue that references this one.