[
https://issues.apache.org/jira/browse/CURATOR-622?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17443262#comment-17443262
]
Jordan Zimmerman edited comment on CURATOR-622 at 11/14/21, 8:53 AM:
---------------------------------------------------------------------
> There is no reason to believe that the instances contending for a latch are
> random in many operational environments
Why not? If n instances contend for a ZNode at around the same time the
sequence number they receive is non-deterministic.
> In the current implementation, the LeaderLatch is operating more as a
> leadership queue than an election system
The current implementation is _fair_. Fairness in locks requires a queue (fair
locks in the JDK do this).
> By adding a random factor into the internal election process...
This will be very hard to do without completely re-writing the LeaderLatch
recipe. The current implementation takes advantage of the sequence numbers when
instances who don't have the lock watch their predecessor - see line 578, the
watch patch is the ZNode path of the previous node.
I'm very much -1 on this change for LeaderLatch. Instead a new recipe should be
written to handle your use case.
was (Author: randgalt):
> There is no reason to believe that the instances contending for a latch are
> random in many operational environments
Why not? If n instances contend for a ZNode at around the same time the
sequence number they receive is non-deterministic.
> In the current implementation, the LeaderLatch is operating more as a
> leadership queue than an election system
The current implementation is _fair_. Fairness in locks requires a queue (fair
locks in the JDK do this).
> By adding a random factor into the internal election process...
This will be very hard to do without completely re-writing the LeaderLatch
recipe. The current implementation takes advantage of the sequence numbers when
instances who don't have the lock watch their predecessor - see line 578, the
watch patch is the ZNode path of the previous node.
I'm very much - 1 on this change for LeaderLatch. Instead a new recipe should
be written to handle your use case.
> Add Randomness to LeaderLatch Elections
> ---------------------------------------
>
> Key: CURATOR-622
> URL: https://issues.apache.org/jira/browse/CURATOR-622
> Project: Apache Curator
> Issue Type: Improvement
> Components: Recipes
> Reporter: Tim Black
> Priority: Major
>
> Currently, LeaderLatch uses EPHEMERAL_SEQUENTIAL nodes, with the next leader
> chosen by the lowest numbered node. In a multi-server environment where each
> server is a participant in multiple elections, the result is that the leader
> will always be the server that has been up the longest.(Or first to be
> restarted during a rolling restart)
> Instead of using sequentially numbered nodes, I propose instead that the node
> number for a new participant be created by adding a random number(From a
> constrained range) to the current leader number.(Defaults to zero) If a node
> with that number exists, repeat until an available node is found. After
> initial node creation, all other aspects of the leader election will remain
> unchanged.
> I have an implementation for this that I am testing locally and will submit a
> PR once the tests are complete.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
