Closed kondaps closed 5 years ago
Hi @kondaps, could you share a bit more information? 700 seems like an awful lot :) Are you running the latest commit? There's been some improvements on this front How long is your retention period? How often do you backup? How many instances are you backing up? Have you deployed this tool more than once in the same account and region?
Hi, We tested with latest commit and it did resolve the issue. Thanks a lot for fixing it already :)
Each deployment is backing up a RDS instance and we are noticing these many times calls and failures due to Rate limit reached with old code. Now we no longer has this issue. I believe the same fix is needed for Aurora tool too.
We are running into API rate limit exceeding errors and per AWS support, the culprit is as below. We need to understand how to correct this issue.
The service team has investigated and has found that the primary API call rate being exceeded is "ListTagsForResource", on 2019-01-09 during a one minute period (at 10:25 UTC), over 700 requests were made, from the following sources:
Boto3/1.7.74 Python/3.6.1 Linux/4.14.88-72.73.amzn1.x86_64 exec-env/AWS_Lambda_python3.6 Botocore/1.10.74 aws-internal/3 aws-sdk-java/1.11.432 Linux/4.9.124-0.1.ac.198.73.329.metal1.x86_64 OpenJDK_64-Bit_Server_VM/25.192-b12 java/1.8.0_192
The vast majority of these are likely to be the Lambda functions mentioned previously.