Endlessh is a great idea that not only blocks the brute force SSH attacks, but also wastes attackers time as a kind of counter-attack. Besides trapping the attackers, I also want to virtualize the Geolocations and other statistics of the sources of attacks. Unfortunately the wonderful C implementation of endlessh only provides text based log, but I do not like the solution that writing extra scripts to parse the log outputs, then exporting the results to a dashboard, because it would introduce extra layers in my current setup and it would depend on the format of the text log file rather than some structured data. Thus I create this golang implementation of endlessh to export Prometheus metrics and a Grafana dashboard to virtualize them.

If you want a dashboard of sources of attacks and do not mind the endlessh server, besides trapping the attackers, does extra things including: translating IP to Geohash, exporting Prometheus metrics, and using more memory (about 10MB), this is the solution for you.

Getting Started

Clone the repo then build from source:

go build .
./endlessh-go

Alternatively, you can use the docker image:

sudo docker run -d shizunge/endlessh-go

It listens to port 2222 by default.

If you want log like the C implementation, you need to set both CLI arguments -logtostderr and -v=1, then the log will go to the stderr. You can set different log destinations via CLI arguments.

Usage

Usage of ./endlessh-go

-alsologtostderr
- log to standard error as well as files
-conn_type string
- Connection type. Possible values are tcp, tcp4, tcp6 (default "tcp")
-enable_prometheus
- Enable prometheus
-geoip_supplier string
- Supplier to obtain Geohash of IPs. Possible values are "off", "ip-api", "freegeoip" (default "off")
-host string
- Listening address (default "0.0.0.0")
-interval_ms int
- Message millisecond delay (default 1000)
-line_length int
- Maximum banner line length (default 32)
-log_backtrace_at value
- when logging hits line file:N, emit a stack trace
-log_dir string
- If non-empty, write log files in this directory
-logtostderr
- log to standard error instead of files
-max_clients int
- Maximum number of clients (default 4096)
-port string
- Listening port (default "2222")
-prometheus_entry string
- Entry point for prometheus (default "metrics")
-prometheus_port string
- The port for prometheus (default "2112")
-stderrthreshold value
- logs at or above this threshold go to stderr
-v value
- log level for V logs
-vmodule value
- comma-separated list of pattern=N settings for file-filtered logging

Metrics

This golang implementation exports the following Prometheus metrics.

Metric	Type	Description
endlessh_client_open_count_total	count	Total number of clients that tried to connect to this host.
endlessh_client_closed_count_total	count	Total number of clients that stopped connecting to this host.
endlessh_sent_bytes_total	count	Total bytes sent to clients that tried to connect to this host.
endlessh_trapped_time_seconds_total	count	Total seconds clients spent on endlessh.
endlessh_client_open_count	count	Number of connections of clients. Labals: `ip`: IP of the client `country`: Country of the IP `location`: Country, Region, and City `geohash`: Geohash of the location
endlessh_client_trapped_time_seconds	count	Seconds a client spends on endlessh. Labals: `ip`: IP of the client

The metrics is off by default, you can turn it via the CLI argument -enable_prometheus.

It listens to port 2112 and entry point is /metrics by default. The port and entry point can be changed via CLI arguments.

The endlessh-go server stores the geohash of attackers as a label on endlessh_client_open_count, which is also off by default. You can turn it on via the CLI argument -geoip_supplier. The endlessh-go uses service from either ip-api or freegeoip, which may enforce a query rate and limit commercial use. Visit their website for their terms and policies.

Dashboard

The dashboard requires Grafana 8.2.

You can import the dashboard from Grafana.com using ID 15156

The IP addresses are clickable and link you to the ARIN database.