1 Answer
- Newest
- Most votes
- Most comments
1
According to Wikipedia a 524 error is a CloudFlare timeout error. Have you tried connecting directly to your instance when this happens?
Without more information this is pretty difficult to troubleshoot. What do the logs say around the time you're having this error? What to the CloudWatch metrics show? Is the instance reaching 100% CPU? If it is a T-series instance s it running out of CPU credits? Is it running out of memory? Is there a spike in network or disk activity?
Relevant content
AWS OFFICIALUpdated a year ago- AWS OFFICIALUpdated 3 years ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated a year ago
@Brettski@AWS Pasted the error response here. https://paste.pics/1b62db3e837c642b529e4c45e19720b9
Below are the details of matrices: (I am not much technical can you please suggest what is causing the issue) https://paste.pics/GOUP2
As a basic website is hosted,so no traffic, I am not sure why high resources is been used.
The smoking gun here is the "CPU credit balance" (very last graph). You're using a T-series instance which has burstable performance; when the CPU is busy it draws down on its credit balance until it gets to zero. At that point you get baseline performance which (depending on the size of the instance) may be very slow - and if other tasks are using the CPU (which it appears they are) then you'll not get the response you require.
Solutions: Use a bigger T series instance (it has a higher credt balance). Choose "unlimited" for your T series instance (that will cost you a little extra but only when you run out of credits). Use a non-T series instance (because they don't have burstable performance - they run at the same speed all the time).
Finally: Look at the process table on your instance (and the logs) to find out what is consuming all of the CPU. Is it web traffic? Is it something else? Fix whatever it is that is causing that; or use the solutions above.