kevlar

Synthetic load generation.

Installation

npm install kevlar

Running

kevlar reads a list of paths from STDIN and resolves them against a target URL (--target) before making HEAD requests.

This means that you can use a static list of paths as a source of load:

cat paths.txt | kevlar -t http://example.com

Or, you can use a dynamic list of paths, in this case processed from a remote syslog stream:

curl -s http://source.example.com | \
grep path | \
perl -pe 's/^.+GET (\/[^\/]+\/[\w\/\.]+).+$/\1/' \
kevlar -t http://example.com

If you see EMFILE errors, kevlar's intended concurrency is being limited by the number of available file descriptors (often 256 by default). To increase the limit, use:

ulimit -n 1024

Notes

kevlar currently makes HEAD requests by default (use -X option to change the method) in order to reduce bandwidth utilization (with the knowledge that in this case the target will perform all necessary work in response).