asterixdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Till Westmann (Code Review)" <>
Subject Change in hyracks[master]: This change fixes the sporadic connection refused errors aft...
Date Sat, 29 Aug 2015 06:12:46 GMT
Till Westmann has posted comments on this change.

Change subject: This change fixes the sporadic connection refused errors after managix starts
up an Asterix instance. This can mostly be seen during asterix-installer tests. The fix basically
lets managix wait and listens through zookeeper for a signal from the cluster c

Patch Set 1:

What confuses me about the design is, why the CC would ask somebody else about the state of
the cluster. Shouldn't the CC know and control the cluster? Let's suppose that we have some
redundancy in the cluster and cluster can survive a node failure. Then it seems that the cluster
controller should know about this and use this knowledge so schedule future jobs. 
Why would the tool that starts and stops the cluster (which is my simplistic view of the role
of managix ...) be the owner of that knowledge?
Do my thoughts make sense or am I missing an important piece of the picture?
(And, yes, I know that that the main goal of the change it to get the tests reliable :) ).

To view, visit
To unsubscribe, visit

Gerrit-MessageType: comment
Gerrit-Change-Id: I0975308e1dc2650d8c2e50d58ff2c696b22eaaa4
Gerrit-PatchSet: 1
Gerrit-Project: hyracks
Gerrit-Branch: master
Gerrit-Owner: abdullah alamoudi <>
Gerrit-Reviewer: Chris Hillery <>
Gerrit-Reviewer: Ian Maxon <>
Gerrit-Reviewer: Jenkins <>
Gerrit-Reviewer: Murtadha Hubail <>
Gerrit-Reviewer: Till Westmann <>
Gerrit-Reviewer: abdullah alamoudi <>
Gerrit-HasComments: No

View raw message