[
https://issues.apache.org/jira/browse/HBASE-15240?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15145151#comment-15145151
]
stack commented on HBASE-15240:
-------------------------------
Messing some more, prefetch needs the above hack where we wait to get our
blocks into cache... otherwise we skip a bunch.... silently. Need some metric
in here, some indicator on how well prefetch is doing. Could we set 'wait on
cache' if prefetching because defaults of not waiting and 50ms seem good as
defaults... though they should be configurable. I can imagine folks finding
50ms too long to wait.
Need to fix the 'full' indicator. Prefetch is erroring out early when full..
and it queues files to read though block cache may be full... FIX.
> Go Big BucketCache Fixes
> ------------------------
>
> Key: HBASE-15240
> URL: https://issues.apache.org/jira/browse/HBASE-15240
> Project: HBase
> Issue Type: Umbrella
> Components: BucketCache
> Reporter: stack
> Assignee: stack
> Attachments: Screen Shot 2016-02-10 at 6.27.26 AM.png
>
>
> Umbrella issue to which we will attach issues that prevent bucketcache going
> big; there's a few.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
