cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <j...@apache.org>
Subject [jira] Commented: (CASSANDRA-2045) Simplify HH to decrease read load when nodes come back
Date Mon, 24 Jan 2011 21:27:43 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-2045?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12986037#action_12986037
] 

Jonathan Ellis commented on CASSANDRA-2045:
-------------------------------------------

It's a tradeoff -- storing the full mutation would be a much bigger hit on writes.  Since
we store hints on other replicas, storing just a "pointer" is almost free.

when we were doing the more expensive HH of storing the hint + row on non-replica nodes, we
saw a lot of cascading failures; this would be a similar bump in extra work done when HH kicks
in, so that scares me.

> Simplify HH to decrease read load when nodes come back
> ------------------------------------------------------
>
>                 Key: CASSANDRA-2045
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2045
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Chris Goffinet
>             Fix For: 0.7.2
>
>
> Currently when HH is enabled, hints are stored, and when a node comes back, we begin
sending that node data. We do a lookup on the local node for the row to send. To help reduce
read load (if a node is offline for long period of time) we should store the data we want
forward the node locally instead. We wouldn't have to do any lookups, just take byte[] and
send to the destination.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message