fix: save app's env when bootstrapping and reloading

[zhongwencool]

If we configure logger and prometheus in cluster-override.conf

log {
  file_handlers {
    default {
      burst_limit {
        enable = true
        max_count = 10000
        window_time = "1s"
      ...
    }
prometheus {
  enable = true
  interval = "151s"
  job_name = "${name}/instance/${name}~${host}"
  mnesia_collector = "enabled"
  push_gateway_server = "http://127.0.0.1:9091"
  vm_dist_collector = "enabled"
  vm_memory_collector = "enabled"
  vm_msacc_collector = "enabled"
  vm_statistics_collector = "enabled"
  vm_system_info_collector = "enabled"
}

These configurations will have no effect. This is because log is an environment variable for kernal applications.
*_collector are the environment variable for the prometheus application.

we need to save those two applications' env after init_load.

close #8895

[zhongwencool]

when we config node.config_files = ["test"]

node {
  name = "emqx@127.0.0.1"
  cookie = "emqxsecretcookie"
  data_dir = "data"
  config_files = ["test"]
}

the config schema check will fail by:

exception throw: {emqx_conf_schema,[#{kind => validation_error,
                                      path => "node.config_files",
                                      reason =>
                                          #{expected_type => <<"[string()]">>},
                                      value => [<<"test">>]}]}
  in function  hocon_tconf:assert_no_error/2 (hocon_tconf.erl, line 1142)
  in call from hocon_tconf:map/4 (hocon_tconf.erl, line 295)
  in call from hocon_cli:get_values/3 (hocon_cli.erl, line 194)
  in call from hocon_cli:multi_get/2 (hocon_cli.erl, line 182)
  in call from escript:run/2 (escript.erl, line 750)
  in call from escript:start/1 (escript.erl, line 277)
  in call from init:start_em/1
  in call from init:do_boot/3

So we change the type to hocon:array(string()).