Sets up an autoscaling cluster of SOCKS proxies to use along with TREVORproxy.

• trevorproxy_serverless uses an SQS queue to control the autoscaling of the proxy cluster
• as long as there is a proxy-intent message in the queue the cluster will stay up.
• when running the cli it adds a message to the queue to signal it needs a running cluster of proxies
• when the cli terminates gracefully it removes the message from the queue
• if the cli is terminated non gracefully, the message will remain in the queue until the message retention period of the queue passes
• while the cli is running it is using a sliding window approach to ensure the proxy intent message will not expire while the cli is running.

the proxy count can be adjusted trough terraform variables.

Dependencies for infra.sh

• aws cli (configured with enough permissions)
• terraform
• docker
• jq

trevorproxy_serverless Installation

$ sudo pipx install --global trevorproxy_serverless

Notes

Cost of one proxy instance running for an hour is 0,012$.

PS: Fargate bills per minute spent so if you spin up 100 servers for 30 minutes it will cost 0,012 * 100 / 2 = 0,60$

PS: Spin up time can take a few minutes because SQS pushes metrics to cloudwatch on an interval of 1 or 5 minues, the following aws regions support 1 minute metrics, so they are recommended: US East (Ohio), EU (Ireland), EU (Stockholm), Asia Pacific (Tokyo)