meminfo collector broken on FreeBSD

[phyber]

Host operating system: output of uname -a

FreeBSD hostname.example.org 12.0-RELEASE-p3 FreeBSD 12.0-RELEASE-p3 GENERIC amd64

node_exporter version: output of node_exporter --version

node_exporter, version 0.18.0 (branch: release-0.18, revision: 1)
  build user:       root
  build date:
  go version:       go1.12.5

node_exporter command line flags

/usr/local/bin/node_exporter \
  --web.listen-address=:9100 \
  --collector.textfile.directory=/var/tmp/node_exporter \
  --collector.ntp

Are you running node_exporter in Docker?

No

What did you do that produced an error?

Run node_exporter with the flags above and had Prometheus scrape metrics.

What did you expect to see?

Swap metrics

What did you see instead?

No swap metrics and the following being logged (long line broken up so people don't have to horizontally scroll as much).

node_exporter[9035]: \
  time="2019-05-14T10:25:49+01:00" \
  level=error \
  msg="ERROR: meminfo collector failed after 0.000133s: couldn't get memory: no such file or directory" \
  source="collector.go:132"

Notes

• Other memory metrics are present, it is only swap that is absent.
• node_exporter appears to be looking for vm.swap_size on FreeBSD instead of vm.swap_total.
  • Commit: 17fee80#diff-d68d35e593b24ae4d0de102cac28f20f
  • PR: Check BSD's mib which accounts for swap size #1149
  • Issue: memory_bsd.go has a sysctl which isn't available in recent DragonflyBSD (not sure about previous one) #1127
• Crude debugging to obtain the error when the mib check happens:

  Mib check err: cannot allocate memory
  Looking for swap mib: vm.swap_size