cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sergio Bossa (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-7816) Duplicate DOWN/UP Events Pushed with Native Protocol
Date Thu, 09 Apr 2015 12:31:13 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-7816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14487290#comment-14487290
] 

Sergio Bossa commented on CASSANDRA-7816:
-----------------------------------------

I find the pending echo check a bit dangerous, as if the echo reply gets swallowed (i.e. because
the connection pool is reset?), the node could stay marked down until the next major state
change. If the objective is to avoid repeated {{Gossiper#realMarkAlive}} calls, wouldn't it
be safer to make the echo callback idempotent? I.e., we could make {{EndpointState#isAlive}}
an {{AtomicBoolean}}, and use CAS in {{EndpointState#markAlive}} and {{EndpointState #markDead}},
returning the CAS result (that is, if it was actually dead and turned alive, and vice-versa):
this could be used by the {{Gossiper}} to conditionally execute {{realMarkAlive}}.

> Duplicate DOWN/UP Events Pushed with Native Protocol
> ----------------------------------------------------
>
>                 Key: CASSANDRA-7816
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-7816
>             Project: Cassandra
>          Issue Type: Bug
>          Components: API
>            Reporter: Michael Penick
>            Assignee: Stefania
>            Priority: Minor
>             Fix For: 2.0.15, 2.1.5
>
>         Attachments: 7816-v2.0.txt, tcpdump_repeating_status_change.txt, trunk-7816.txt
>
>
> Added "MOVED_NODE" as a possible type of topology change and also specified that it is
possible to receive the same event multiple times.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message