Factor | EBS backed AMI's | S3 backed AMI's |
---|---|---|
Storage Size | 1TB | 10GB |
Instances can be stopped | Yes | No |
Boot Time | Faster (~1min) | Slower (~5mins) |
Local Instance Store | Not Available | Available |
Data Persistence | EBS attached as volume (persist) | S3 is the root partition (doesn't persist) |
AMI Creation | Easier and can be done using API | No API available for Linux |
Charges | Volume Storage + storage usage + AMI Storage + Instance usage | AMI Storage + instance usage |
Customized AMI Storage Charges | Lower (charged only for the changes) | higher (full storage charge) |
Instance Usage Charge | No charge for stopped instances. Charge full instance hour for every transition from stopped to running state (.10 for full hour running, .30 for restarting twice in a hour) | Not Applicable |
Amazon EC2 Limitations:
- 20 running instances
- Overall instance limit = 4xthe running instance limit
- 100 Amazon EBS Volumes or 20TB total volume storage
- 100 Security Groups
What does this mean?
ReplyDeleteSorry, I cannot edit using this silly comments system. My question is -- which one do you recommend in which scenario? Just listing the table is fine, but a bit of view point would be more useful.
DeleteAlso, the "20 running instances" means per account??? That's useless. What if I have more than 20 websites, each running in its instance?