ssb-proc

ssb-proc

add process stats to sbot status

example

> sbot plugins.install ssb-proc
> # now, restart sbot...

> sbot status

output: (snipped at ...)

{
  ...
  "proc": {
    "pid": 16873,
    "uptime": 170.073,
    "memoryUsage": {
      "rss": 187174912,
      "heapTotal": 75300864,
      "heapUsed": 68445192,
      "external": 21983959
    },
    "cpuUsage": {
      "user": 4679999,
      "system": 496666
    },
    "blockyness": 0.96,
    "loadavg": [
      0.37158203125,
      0.5224609375,
      0.45458984375
    ],
    "io": {
      "rchar": 78338626,
      "wchar": 796716,
      "syscr": 7711,
      "syscw": 1999,
      "read_bytes": 0,
      "write_bytes": 901120,
      "cancelled_write_bytes": 0
    },
    "net": {
      "lo:": {
        "rx": 740057,
        "tx": 740057
      },
      "wlp2s0:": {
        "rx": 214996,
        "tx": 111649
      }
    }
  }
}

status.proc

under the proc heading, are the following measurements

pid

the process identifier of the node process.

uptime

length of time in seconds that the current sbot process has been running.

memoryUsage

node.js process memory usage

blockyness

A the ratio of setIntervals that have fired in the last second. If this is not near 1 something must be hogging the CPU, possibly heavy loops.

loadavg

average cpu usage in the last minute, 5 minutes and 15 minutes.

io

output of /proc/{pid}/io

rchar and wchar is reads and write calls to the os, including cached reads. read_bytes and write_bytes is the actual bytes that have been read or written to disk, by this process since it started.

net

Bytes received (rx) and transmitted per network interface, taken from /proc/net/dev

Unlike the other measurements, This applies to the system as a whole and not the specific process. I would like to make it per process - any suggestions or PR most appreciated!

License

MIT