aws-signing-proxy is a proxy service, written in go, for automatically signing requests made to AWS endpoints. It leverages the aws-sdk-go library to sign requests to arbitrary URLs in AWS. I wrote it to connect a kibana instance to an AWS Elasticsearch cluster using an IAM role instead of hard-coding IPs in the access policy. Other uses may exist.
Docker image: https://hub.docker.com/r/cllunsford/aws-signing-proxy/
export AWS_ACCESS_KEY_ID=<xxx>
export AWS_SECRET_ACCESS_KEY=<xxx>
export AWS_REGION=<xxx>
export AWS_PROFILE=<xxx> # Optional
./aws-signing-proxy -target https://search-my-cluster.us-west-2.es.amazonaws.com [-port 8080] [-service es]
Flags
General:
-target
- AWS service to send requests to. Required.-port
- Port for the proxy to LISTEN on (will forward to whatever port you specify in target), default:8080
.-service
- The AWS service type you are sending to, default:es
. This is required for the signing process.
HTTP Connection Tuning:
-flush-interval
- ReverseProxy FlushInterval, default:0
-idle-conn-timeout
- Transport Idle Connection Timeout, default:90s
-dial-timeout
- Transport Dial Timeout, default:30s
AWS credentials are looked up in the following order:
- Environment variable accessible to the app
- The
~/.aws/credentials
config file's default profile - Any IAM instance profile role assigned to the instance.
If you have go installed, you can build and install the binary natively:
go install
The Makefile is used for the production build in travis. It builds the binary in a docker container. After installing docker:
make gobuild # creates the binary
make dockbuild # creates a docker image with the binary
or
make build # does both
If you see:
{"message":"Signature expired: 20160415T172935Z is now earlier than 20160415T174424Z (20160415T174924Z - 5 min.)"}
verify that the clock/time is in sync on the proxy host.
For AWS Elasticsearch, the built-in kibana populates the .kibana index. If you see:
ClusterBlockException[blocked by: [FORBIDDEN/8/index write (api)];]
try changing the kibana index setting to use a different index. The marcbachmann/kibana4 docker image allows you to change this easily by setting the KIBANA_INDEX
environment variable.
MIT 2018 (c) Chris Lunsford