cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Pavel Yaskevich (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-6689) Partially Off Heap Memtables
Date Tue, 25 Mar 2014 00:59:42 GMT


Pavel Yaskevich commented on CASSANDRA-6689:

bq. One last change I did not include was your suppression of errors from calling the direct
cleaner: IMO this is dangerous, and the user should immediately find out that we do not have
full support for off-heap functionality by crashing, so that they can reconfigure to use the
on-heap allocator only. If we cannot clean, we are at risk of completely trashing the file
system page cache through competition with our direct buffers. Better to fail badly ASAP.

First of all it would not "fail badly ASAP" because that cause ClassCast or NoSuchMethod exception,
which would just get logged but the executing thread. It's better to error to the log with
something useful e.g. that the cleanup is not going to happen rather then fails with ClassCastException
which doesn't say much to the users the same way we do for FileUtils.

[~jbellis] Can you remind us what is the general policy on the crashing? :)

> Partially Off Heap Memtables
> ----------------------------
>                 Key: CASSANDRA-6689
>                 URL:
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core
>            Reporter: Benedict
>            Assignee: Benedict
>              Labels: performance
>             Fix For: 2.1 beta2
>         Attachments: CASSANDRA-6689-final-changes.patch, CASSANDRA-6689-small-changes.patch
> Move the contents of ByteBuffers off-heap for records written to a memtable.
> (See comments for details)

This message was sent by Atlassian JIRA

View raw message