- 最新
- 投票最多
- 评论最多
The thing that sticks out at me here is the CPU credit balance
graph. Note that it goes to zero at 23:15. This is going to cause the performance of the instance to drop to baseline which may be impacting your availability.
You can solve this by choosing "Unlimited" mode for the instance. Whenver the CPU credit balance goes to zero the performance will not be impacted; but you will be charge for that service. If this is happening regularly a better option would be to use a different instance type such as the M, R or C instance types.
The best recommendation would be to create a support ticket and they can help with this.
Hello,
In addition to Brettski's answer of switching to "unlimited mode" I would also advise to check memory usage of the instance, from our documentation [1] common reasons for StatusCheckFailed_Instance are as following:
- Failed system status checks
- Incorrect networking or startup configuration
- Exhausted memory
- Corrupted file system
- Incompatible kernel
Based off the symptoms you described and your approach to fix it sounds a lot like a memory issue, see docs [2] that states Reboot the instance to return it to a non-impaired status. The problem will probably occur again unless you change the instance type
As you know that EC2 does not natively monitor memory metrics, we have 2 approaches to configure this manually adding them below with links:
In addition to the resources above also adding this third party link [3] which does a good job of summarizing and configuring the aws agent to log memory utilization.
I do hope the above resources help, feel free to comment on this thread if there is anything else we can assist with.