2개 답변
- 최신
- 최다 투표
- 가장 많은 댓글
0
Don't know if you've sorted out your problems yet, but I'd suspect that your jobs are running out of their I/O allowance.
Every EC2 instance - and, in particular, their attached EBS storage - has an I/O budget. If you consume I/O quickly, you eventually run out of your 'burst allowance', and get throttled back.
You can verify that by using the CloudWatch stats to monitor the burst allowance for your EBS devices.
답변함 5년 전
0
I didnt know EBS had burst, so thank you for that. You are most likely right. It would make sense why a new volume would be back to the original speed.
Thank you
답변함 5년 전