bowenliang123 commented on code in PR #5182:
URL: https://github.com/apache/kyuubi/pull/5182#discussion_r1306477272


##########
kyuubi-common/src/main/scala/org/apache/kyuubi/config/KyuubiConf.scala:
##########
@@ -2614,6 +2614,17 @@ object KyuubiConf {
       .toSequence()
       .createWithDefault(Nil)
 
+  val SERVER_LIMIT_CONNECTIONS_USER_LIMITED_LIST: ConfigEntry[Seq[String]] =
+    buildConf("kyuubi.server.limit.connections.user.limited.list")
+      .doc("The user in the limited list will be limited to connect to kyuubi 
server, " +
+        "if the user has configured both user.unlimited.list and 
user.limited.list, " +
+        "the priority of the latter is higher.")
+      .version("1.8.0")
+      .serverOnly
+      .stringConf
+      .toSequence()

Review Comment:
   > I had an offline discussion with @turboFei, and we reached a consensus 
that:
   > 
   > 1. The current implementation is **deny** semantic, which is a good 
feature but **limited** is not a properly name
   > 2. You could either change the configuration name to `...deny.user.list`, 
or keep the `...limited.user.list` and change the implementation as I commented 
above [[KYUUBI #4540] [Improvement] Support to limit some user to make 
connection. #5182 
(comment)](https://github.com/apache/kyuubi/pull/5182#discussion_r1304545505), 
I think the latter approach is simple, and we can implement the **deny** in 
another PR.
   
   I have similar concerns in these variations. I was confused by the **deny** 
and **limited**.
   As for me, a limited user list is not the contrast to an unlimited user, and 
only a feature for  deny users list is necessary.
   What's more, I would like to see when the server rejects user connection , 
this feature should be part of authentication, such as User Filter in LDAP auth 
implementation,  but not an injected mechanism in such Limiter. Or did I miss 
any information for better understanding?
   
   I would suggest, 
   1. Add detailed and summarised design and key points to this PR description, 
as a simple link to the original issue is not enough
   2. Add constructive discussion in proposed change and the reason for 
possible approaches
   
   I'm fine if we accept this feature,  when we have a mutual understanding of 
enough information.



-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to