[erlang-questions] Lots of epmd in pod

bengt cean.ebengt@REDACTED
Fri Aug 9 08:42:29 CEST 2019


There is an Erlang release(*) running in a Kubernetes pod. When I use the start script (even if it is only to see the arguments) I get more epmd processes(**).
Is that just me? A known problem that can be ignored? Worth investigation?

Best Wishes,

The Docker file has:
erlang: and alpine:3.9 (one for build and one for container)

Build command:
RUN	rebar3 as prod release

The rebar.config:
{relx, [{release, { 'ercdf', "1.3.7" }, [ercdf, recon, observer_cli]},
{dev_mode, true},
{include_erts, false},
{extended_start_script, true}]

Example after having kubectl exec'd:
 # ps | grep epmd | wc
       46       186      1326
 #  /opt/ercdf/bin/ercdf remote_console
Erlang/OTP 21 [erts-] [source] [64-bit] [smp:8:8] [ds:8:8:10] [async-threads:1] [hipe]

Eshell V10.3.5.4  (abort with ^G)
User switch command
 --> q
 # ps | grep epmd | wc
       58       234      1662

Sample of the set of processes that grows
 2136 root      0:00 [erl_child_setup]
 2166 root      0:00 [inet_gethost]
 2226 root      0:00 [epmd]
 2227 root      0:00 [epmd]

More information about the erlang-questions mailing list