pkuwm commented on a change in pull request #838: Make ConfigAccessor and
ZkUtil realm-aware
URL: https://github.com/apache/helix/pull/838#discussion_r386792509
##########
File path: helix-core/src/main/java/org/apache/helix/manager/zk/ZKUtil.java
##########
@@ -92,15 +97,17 @@ public static boolean isClusterSetup(String clusterName,
HelixZkClient zkClient)
requiredPaths.add(PropertyPathBuilder.controllerHistory(clusterName));
boolean isValid = true;
- BaseDataAccessor<Object> baseAccessor = new
ZkBaseDataAccessor<Object>(zkClient);
- boolean[] ret = baseAccessor.exists(requiredPaths, 0);
+ boolean[] ret = new boolean[requiredPaths.size()];
+ for (int i = 0; i < requiredPaths.size(); i++) {
+ ret[i] = zkClient.exists(requiredPaths.get(i));
Review comment:
I am still not very convinced we change the behavior. I understand the
functionality may be the same. But if one call is blocked because of connection
and it has to retry, other calls could be blocked. I am not sure how critical
it is for here. Maybe Junkai can give an opinion. Since we are also integrating
baseDataAccessor, can we just keep this original behavior by calling base data
accessor?
----------------------------------------------------------------
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:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
