https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/TroubleshootingInstancesStopping.html
Cloudwatch -> Metrics -> Explorer
Explorer -> EC2
Explorer -> Elastic Block Store
From -> InstanceId: i-000528b883bf29b49
https://aws.amazon.com/premiumsupport/knowledge-center/ec2-linux-resource-over-utilization/
sudo ufw status
sudo ufw disable
System status checks System reachability check passed
Instance status checks Instance reachability check failed
Note: A stop and start isn't equivalent to a reboot. A start is required to migrate the instance to healthy hardware.
The instance must be migrated to a new, healthy host by stopping and starting the instance. You can wait for Amazon EC2 to perform the stop and start of your instance. Or, you can manually stop and start the instance to migrate it to a new, healthy host.
When the instance is in the running state, choose Actions, Monitor and troubleshoot, Get system log.
https://forums.aws.amazon.com/thread.jspa?threadID=84461
By default EC2 instances are not redundant, Redundancy is achieved through using our other services such as Auto-scaling and Elastic lad balancing where you can mitigate around instance failure and make your application resilient of the components that it's running on.
You have to construct your server stack in a way that you don't care of the entire server goes away. In other words, keep your data in non-instance storage (EBS volumes, RDS, SimpleDB, S3). You have to have the mentality that your "instance" can go away at any time and it won't bother you, because you can click once and start up another one.
https://aws.amazon.com/ec2/instance-types/
https://towardsdatascience.com/choosing-the-right-gpu-for-deep-learning-on-aws-d69c157d8c86
No comments:
Post a Comment