[🐛 BUG]: rr workers/reset commands don't respect default config values

[r4m-alexd]

No duplicates 🥲.

• I have searched for a similar issue in our bug tracker and didn't find any solutions.

What happened?

A bug happened!

Version (rr --version)

rr version 2024.1.1 (build time: 2024-05-02T16:37:17+0000, go1.22.2), OS: linux, arch: amd64

How to reproduce the issue?

1. Create a config file .rr.test.yaml:

version: '3'

server:
  command: "php8.1 psr-worker.php"

http:
  address: 0.0.0.0:8080

rpc:
  listen: tcp://0.0.0.0:${RR_RPC_PORT:-6081}

logs:
  level: debug
  mode: development

2. Start rr with the command ./rr -c .rr.test.yaml serve:

2024-05-07T14:34:38+0000        DEBUG   rpc             plugin was started      {"address": "tcp://0.0.0.0:6081", "list of the plugins with RPC methods:": ["lock", "resetter", "app", "informer"]}
2024-05-07T14:34:38+0000        DEBUG   server          worker is allocated     {"pid": 13914, "max_execs": 0, "internal_event_name": "EventWorkerConstruct"}
2024-05-07T14:34:38+0000        DEBUG   server          worker is allocated     {"pid": 13916, "max_execs": 0, "internal_event_name": "EventWorkerConstruct"}
2024-05-07T14:34:38+0000        DEBUG   server          worker is allocated     {"pid": 13915, "max_execs": 0, "internal_event_name": "EventWorkerConstruct"}
2024-05-07T14:34:38+0000        DEBUG   server          worker is allocated     {"pid": 13913, "max_execs": 0, "internal_event_name": "EventWorkerConstruct"}
[INFO] RoadRunner server started; version: 2024.1.1, buildtime: 2024-05-02T16:37:17+0000
[INFO] sdnotify: not notified
2024-05-07T14:34:38+0000        DEBUG   http            http server was started {"address": "0.0.0.0:8080"}

3. Try to get info about workers using the command ./rr -c .rr.test.yaml workers

dial tcp 0.0.0.0:0: connect: connection refused

This one doesn't work either :

$ RR_RPC_PORT=6081 ./rr -c .rr.test.yaml workers
dial tcp 0.0.0.0:0: connect: connection refused

Only this command works OK:

$ ./rr -c .rr.env.yaml -o "rpc.listen=tcp://0.0.0.0:6081"  workers
Workers of [http]:
+---------+-----------+---------+---------+---------+--------------------+
|   PID   |  STATUS   |  EXECS  | MEMORY  |  CPU%   |      CREATED       |
+---------+-----------+---------+---------+---------+--------------------+
|   13913 | ready     |       0 | 50 MB   |    0.20 | 14 seconds ago     |
|   13914 | ready     |       0 | 49 MB   |    0.26 | 14 seconds ago     |
|   13915 | ready     |       0 | 49 MB   |    0.20 | 14 seconds ago     |
|   13916 | ready     |       0 | 49 MB   |    0.20 | 14 seconds ago     |
+---------+-----------+---------+---------+---------+--------------------+

4. Same problem with the reset command:

$ ./rr -c .rr.test.yaml reset
dial tcp 0.0.0.0:0: connect: connection refused

Relevant log output

No response

[rustatian]

Hey @r4m-alexd 👋
rr reset command should be used with a host and port, w/o env variables, because RR doesn't use the config parser for these commands. Please try to specify a port directly and retry the command.

EDIT:
Hm, I was wrong, we use the envs parser. Looks like this is the real bug, will be fixed in the next bugfix release, thanks.

[rustatian]

./rr -c .rr.env.yaml -o "rpc.listen=tcp://0.0.0.0:6081" workers this command works because you override the host and port with the real values.

[rustatian]

@r4m-alexd Fixed