cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stefania (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-7807) Push notification when tracing completes for an operation
Date Tue, 24 Mar 2015 02:59:53 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-7807?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14377179#comment-14377179
] 

Stefania commented on CASSANDRA-7807:
-------------------------------------

Sorry to add more but there is something I totally missed yesterday.

I didn't realize {{Server.ConnectionTracker}} is already filtering by event type and sends
events only to channels that have registered for a specific type, see the {{groups}} map.
In view of this, IMHO it would be sufficient to send to all clients who registered for tracing
events (same as the other events) and then the client looks at the trace uuid to see if it
is of interest. This would make the implementation much simpler and in line with the other
client notifications, we just register {{Server.EventNotifier}} as an observer of {{Tracing}},
which is then passed to all {{TraceState}} instances, either by adding a new interface or
by re-using {{ProgressListener}}.

[~thobbs], [~iamaleksey] do you think it would be OK to pass this complexity to the clients
or do we really have to ensure clients only receive trace notifications for trace messages
they originated?


> Push notification when tracing completes for an operation
> ---------------------------------------------------------
>
>                 Key: CASSANDRA-7807
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-7807
>             Project: Cassandra
>          Issue Type: Sub-task
>          Components: Core
>            Reporter: Tyler Hobbs
>            Assignee: Robert Stupp
>            Priority: Minor
>              Labels: client-impacting, protocolv4
>             Fix For: 3.0
>
>         Attachments: 7807.txt
>
>
> Tracing is an asynchronous operation, and drivers currently poll to determine when the
trace is complete (in a loop with sleeps).  Instead, the server could push a notification
to the driver when the trace completes.
> I'm guessing that most of the work for this will be around pushing notifications to a
single connection instead of all connections that have registered listeners for a particular
event type.



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

Mime
View raw message