emqx/apps/emqx_resource

emqx_resource

The emqx_resource is an application that manages configuration specs and runtime states for components that need to be configured and manipulated from the emqx-dashboard.

It is intended to be used by resources, actions, acl, auth, backend_logics and more.

It reads the configuration spec from *.spec (in HOCON format) and provide APIs for creating, updating and destroying resource instances among all nodes in the cluster.

It handles the problem like storing the configs and runtime states for both resource and resource instances, and how porting them between different emqx_resource versions.

It may maintain the config and data in JSON or HOCON files in data/ dir.

After restarting the emqx_resource, it re-creates all the resource instances.

There can be foreign references between resource instances via resource-id. So they may find each other via this Id.

Try it out

$ ./demo.sh
Eshell V11.1.8  (abort with ^G)
1> == the demo log tracer <<"log_tracer_clientid_shawn">> started.
config: #{<<"config">> =>
            #{<<"bulk">> => <<"10KB">>,<<"cache_log_dir">> => <<"/tmp">>,
                <<"condition">> => #{<<"clientid">> => <<"abc">>},
                <<"level">> => <<"debug">>},
        <<"id">> => <<"log_tracer_clientid_shawn">>,
        <<"resource_type">> => <<"log_tracer">>}
1> emqx_resource_instance:health_check(<<"log_tracer_clientid_shawn">>).
== the demo log tracer <<"log_tracer_clientid_shawn">> is working well
state: #{health_checked => 1,logger_handler_id => abc}
ok

2> emqx_resource_instance:health_check(<<"log_tracer_clientid_shawn">>).
== the demo log tracer <<"log_tracer_clientid_shawn">> is working well
state: #{health_checked => 2,logger_handler_id => abc}
ok

3> emqx_resource_instance:query(<<"log_tracer_clientid_shawn">>, get_log).
== the demo log tracer <<"log_tracer_clientid_shawn">> received request: get_log
state: #{health_checked => 2,logger_handler_id => abc}
"this is a demo log messages..."

4> emqx_resource_instance:remove(<<"log_tracer_clientid_shawn">>).
== the demo log tracer <<"log_tracer_clientid_shawn">> stopped.
state: #{health_checked => 0,logger_handler_id => abc}
ok

5> emqx_resource_instance:query(<<"log_tracer_clientid_shawn">>, get_log).
** exception error: {get_instance,{<<"log_tracer_clientid_shawn">>,not_found}}