zookeeper-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rajini Sivaram (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ZOOKEEPER-3100) ZooKeeper client times out due to random choice of resolved addresses
Date Mon, 23 Jul 2018 16:13:00 GMT

    [ https://issues.apache.org/jira/browse/ZOOKEEPER-3100?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16553039#comment-16553039

Rajini Sivaram commented on ZOOKEEPER-3100:

[~andorm] In the failing tests, ZooKeeper was being started on `` explicitly ([https://github.com/apache/kafka/blob/trunk/core/src/test/scala/unit/kafka/zk/EmbeddedZookeeper.scala).]
Clients were connecting using `localhost`. We are fixing this for now in the tests by changing
the clients to use ``, but it used to work fine before upgrading ZK clients.

> ZooKeeper client times out due to random choice of resolved addresses
> ---------------------------------------------------------------------
>                 Key: ZOOKEEPER-3100
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-3100
>             Project: ZooKeeper
>          Issue Type: Bug
>          Components: java client
>    Affects Versions: 3.4.13
>            Reporter: Rajini Sivaram
>            Priority: Major
> The changes to ZooKeeper clients to re-resolve hosts made under ZOOKEEPER-2184 results
in delays when only a subset of the addresses that a host resolves to are actually reachable.
This can result in connection timeouts on the client.
> For example, when running tests with a single ZooKeeper server accepting connections
on on a host that has both IPv4 and IPv6, we have seen connection timeouts in tests
if client connects using `localhost` rather than ``. ZooKeeper client resolves `localhost`
to both the IPv4 and IPv6 addresses and chooses a random one. If IPv6 was chosen, a fixed
one second backoff is applied before retry since there is only one hostname specified. After
backoff, 'localhost' is resolved again and a random address chosen, which could also be the
unconnectable IPv6 address.
> For the list of host names specified for connection, the clients do round-robin without
backoffs until connections to all hostnames are attempted. Can we also do the same for addresses
that each of the hosts resolves to, so that backoffs are only applied after connection to
each address is attempted once and every address is connected to once using round-robin rather
than random selection? This will avoid delays in cases where at least one address can be connected

This message was sent by Atlassian JIRA

View raw message