A bot to measure latency between homeservers, as perceived by users.
Assuming Go 1.9 is already installed on your PATH (also possible with vgo
):
# Get it
go get github.com/silkeh/matrix-monitor-bot/cmd/...
# Configure it (edit monitor-bot.yaml to meet your needs)
cp config.sample.yaml monitor-bot.yaml
# Run it
bin/monitor_bot
The steps are almost the same as above. The only difference is that gb build
will not work, so instead use the following lines:
go build -o bin/monitor_bot ./src/github.com/silkeh/matrix-monitor-bot/cmd/monitor_bot/
/path/to/matrix-monitor-bot
should always be pointed to a folder that has your monitor-bot.yaml
file in it. If the config
file does not exist, one will be created for you (and promptly not work because it doesn't have a valid config). A folder
named logs
will also be created here (assuming you use the default configuration).
From Docker Hub:
docker run -p 8080:8080 -v /path/to/matrix-monitor-bot:/data silkeh/matrix-monitor-bot
Build the image yourself:
git clone https://github.com/silkeh/matrix-monitor-bot
cd matrix-monitor-bot
docker build -t matrix-monitor-bot .
docker run -p 8080:8080 -v /path/to/matrix-monitor-bot:/data matrix-monitor-bot
If metrics are enabled in your config, matrix-monitor-bot will serve up metrics for scraping by Prometheus. Every metric that is exported is a Histogram metric. The following metrics are exported:
monbot_ping_send_delay_seconds
- Number of seconds for the origin to send a ping to their homeservermonbot_ping_receive_delay_seconds
- Number of seconds for a bot to receive a pingmonbot_ping_time_seconds
- Total number of seconds for a ping to go through (from the remote bot to the receiver)
TODO: This section
- How the bot measures things
- What the Prometheus metrics are
- Why the bot uses m.room.message and not a custom event
- Why the bot uses messages for pongs instead of read receipts
- Why the display name gets overwritten and how it is used
- Why someone should run this on their server