[
https://issues.apache.org/jira/browse/KAFKA-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14346083#comment-14346083
]
Geoffrey Anderson commented on KAFKA-902:
-----------------------------------------
Thanks for the feedback.
Is there any reason to keep 0 -> 0 behavior? In other words, is there any harm
to making 0 less magical and perturbing it just like any other value?
Also, I like the idea of scaling the upper bound on jitter with backoff_ms,
though it seems like we still want some amount of jitter even if backoff_ms is
small (in your example, if backoff_ms < 5, then jitter is always 0)
In that case, we might want to make sure that the jitter can be non-zero with
something like max(3, min(20, 0.2 * backoff_ms))
But then we end up with a couple more semi-arbitrary parameters - a scaling
parameter and a hard minimum.
> Randomize backoff on the clients for metadata requests
> ------------------------------------------------------
>
> Key: KAFKA-902
> URL: https://issues.apache.org/jira/browse/KAFKA-902
> Project: Kafka
> Issue Type: Bug
> Components: clients
> Affects Versions: 0.8.0
> Reporter: Neha Narkhede
> Assignee: Geoffrey Anderson
> Priority: Critical
> Labels: newbie
> Attachments: KAFKA-902.patch
>
>
> If a Kafka broker dies and there are a large number of clients talking to the
> Kafka cluster, each of the clients can end up shooting metadata requests at
> around the same time. It is better to randomize the backoff on the clients so
> the metadata requests are more evenly spread out
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
