Syslog proxy that receives syslogs from Fastly, and sends them to Loggly.
var lylog = require('lylog')
var proxy = lylog({
token: 'loggly user token here',
// optional options described below
})
proxy.listen(10514)
// now go tell fastly to send syslogs to logs.mysite.com:10514,
// and they'll be proxied to loggly
Or in bash:
# same as the above JavaScript
lylog \
--token 'loggly user token here' \
--listen 10514
$ lylog -h
lylog [options]
Options:
-l PORT, --listen=PORT Port that lylog listens on. Required.
-t TOKEN, --token=TOKEN Loggly customer token. Required.
-p PORT, --port=PORT Port to send Loggly messages to (default=514)
-H HOST, --host=HOST Hostname to send Loggly messages to
(default=logs-01.loggly.com)
-e, --echo Print logs to stdout as well as sending them to
loggly
-w LIST, --whitelist=LIST Whitespace-separated list of IPv4 blocks to
accept. Defaults to the known list of Fastly IPv4
address blocks as of lylog publish date (see
"fastly-ips.txt" in lylog root). Set to "ALLOW"
to allow connections from any client (not
recommended!)
-h, --help Display this help
token String, Required. Your loggly customer token.host String, Default='logs-01.loggly.com'. The hostname on
loggly where logs are sent.port Number, Default=514. Port on loggly to send logs.whitelist Array(String) or false. Array of Fastly IPv4 blocks
to accept connections from. (Note: without this, your server is an
open proxy that will spam your loggly account!)echo Boolean, default=false. Print logs to stdout as well as
sending to loggly. Useful if you want to maintain a local copy as
well.onerror Function, Optional. Call this function when
upstream connections to loggly experience an error writing.minFree Number, Optional, Default=16. Minimum number of free
connections to loggly before reaping will occur.freeReapRatio Number, Optional, Default=0.5. Minimum ratio of
free loggly connections to total connections before reaping will
occur. Set to 1 to never reap connections (not recommended!)Outbound connections to loggly are created as needed, and then kept
alive for future use. If the free list of connections is greater
than 16 (minFree config), and more than 50% of the total pool of
connections (freeReapRatio config), then the bottom 50% of the free pool
will be reaped. This prevents spikes from causing an abnormal number
of connections to be open indefinitely.