Distributed logger
npm install hypercore-logger
const Logger = require('hypercore-logger')
// pass backing hypercore
const log = new Logger(core)
// add a log message
await log.log({ hello: 'world' })
Setup swarming with Hyperswarm on the core and somewhere else
const log = new Logger(core)
for await (const { timestamp, stats, message } of log.tail()) {
console.log(timestamp, stats, message)
}
Loggerconst log = new Logger(core)Create a new logger backed by a Hypercore (core).
log.discoveryKeyThe discoverKey of the Hypercore backing the log.
log.keyThe key of the Hypercore backing the log.
await log.ready()Wait for the log to fully open.
await log.log(...msg)Log the msg arguments as a string representation intended for debugging, like console.log()'s output.
In addition to the string representation, metadata is automatically added to each log. Each appended to the backing Hypercore will look like:
{
timestamp: Date.now(),
stats: Stats,
subsystem: log.subsystem,
message: '' // String representation of msg
}
Stats includes:
stats: {
cpus: number, // Number of avaliable cpus
cpu: number, // Precent of CPU usage
cpuThread: number, // Percent of thread usage
cpuDelay: number, // Event loop delay the last ~5s
rss: integer, // Resident Set Size
heapUsed: integer, // Amount of heap used in js engine
heapTotal: integer, // Total size of heap in js engine
external: integer, // Memory usage of C++ objects bound to JavaScript objects
}
The block is encoded via the Hyperschema schema @logger/entry.
const stream = log.tail(opts)Returns a live Readable Stream of the contents of the log. Acts like tail -f
on a file, showing new data as it's appended.
Example:
for await (const { timestamp, stats, message } of log.tail()) {
console.log(timestamp, stats, message)
}
await logger.find(opts)Find start and end blocks to apply filters
opts.gte: Start at the first entry with a timestamp >= gte
opts.gt: Start at the first entry with a timestamp > gt
opts.lte: End before the first entry with timestamp > lte
opts.lt: End before the first entry with timestamp >= lt
await log.close()Fully close the log and the Hypercore backing it.
Comes with a CLI tool for tailing a hypercore-logger.
npm install -g hypercore-logger
hypercore-logger --key <key>
Usage:
hypercore-logger [flags]
Flags:
--key, -k <key> log core key
--peer, -p <key> noise key of peer
--storage, -s <dir> where to store the core, defaults to /tmp/hypercore-logger
--help|-h Show help
Apache-2.0