wilfred-s commented on a change in pull request #17:
URL:
https://github.com/apache/incubator-yunikorn-scheduler-interface/pull/17#discussion_r432530581
##########
File path: scheduler-interface-spec.md
##########
@@ -792,6 +792,31 @@ message ForgotAllocation {
}
```
+#### Event Cache
+
+The Event Cache is a SchedulerPlugin that exposes events about scheduler
objects aiming to help the end user to
+see these events from the shim side. An event is sent to the shim side through
the callback in a form of an `EventMessage`.
+
+```protobuf
+message EventMessage {
+ enum Type {
+ REQUEST = 0;
+ APP = 1;
+ NODE = 2;
+ QUEUE = 3;
+ }
+
+ // the type of the object associated with the event
+ Type type = 1;
+ // ID of the object associated with the event
+ string ID = 2;
Review comment:
Name of the object is ambiguous, could be just "application", it does
not really show that it is the ID to identify just one object. It is the ID and
an ID does not have to be unique. It refers to an object type with the ID,
`objectID` covers it probably better.
It also stays in line with the other objects (node and app) that we are
using in other messages. The only one outside of that is the queue which now
has just a name when passed with the but in the event should have the full
path. `objectID` again covers it correctly because it is unique for that type.
Can you also explain the reason behind adding `group` to the event? Is that
the `parent` object the event can get aggregated into? If so group did not make
that clear, I had to deduce that from the reference given for the different
types.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
