This is a collection of Amazon EC2 system tools for Linux.

The implementation is a single Python script with no package dependencies except the standard library.

It runs on both Python 2.x and 3.x.

To authenticate, either provide machines with an IAM role or set environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY. If an IAM role is provided, usually no further configuration is needed.

The script automatically pulls down region and instance id, unless provided in the environment variables AWS_REGION and AWS_INSTANCE_ID.

The metrics command posts system metrics to AWS CloudWatch <https://aws.amazon.com/cloudwatch/>_:

$ python aws-system-tools.py metrics /tmp/stats

Metrics:

• MemoryUtilization
• DiskSpaceUtilization
• LoadAverage
• NetworkConnections

If the statfile argument is provided, an additional set of metrics are reported:

• User: Normal processes executing in user mode
• Nice: Niced processes executing in user mode
• System: Processes executing in kernel mode
• Blocked: Waiting for I/O to complete
• Irq: Servicing interrupts
• SoftIrq: Servicing softirqs
• Steal: Virtual CPU steal

Each metric is reported with respect to both the instance id and the image id, the latter providing aggregate numbers for all instances running that image (e.g. auto-scale groups).

The snapshot command requests an EBS snapshot of those volumes currently attached logically to the machine.

$ python aws-system-tools.py snapshot

The rds-log-sync command synchronizes log files from an RDS instance to a local directory.

$ python aws-system-tools.py rds-log-sync <db-instance-id> <path>

The path argument is optional and defaults to the working directory.

If --metrics is given then the script submits CloudWatch metrics with statistics on how frequently the script has run (without failure) and how much it has downloaded (number of requests, bytes and throughput).