[
https://issues.apache.org/jira/browse/HBASE-5074?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13204336#comment-13204336
]
Phabricator commented on HBASE-5074:
------------------------------------
stack has commented on the revision "[jira] [HBASE-5074] Support checksums in
HBase block cache".
Took a look at a little piece of the patch. It looks great.
INLINE COMMENTS
src/main/java/org/apache/hadoop/hbase/HConstants.java:601 It looks like this
feature will be on by default. Good.
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:1 Should this
class be in an fs package rather than in util?
Nit. HFileSystem seems overly generic. Should it be HBaseFileSystem?
Out of interest, is there a performance penalty that you know of going via
FilterFileSystem?
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:40 How would this
happen? We'd look at the path for the object and do a different fs in here
based off that?
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:49 Won't the
master use this fs too?
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:50 configuration
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:74 cool
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:107 Who would
want this? Can we shut it down?
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:112 Its not the
'default' fs, it IS the fs?
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:167 cool
src/main/java/org/apache/hadoop/hbase/util/HFileSystem.java:172 So we'll have
nonrecursive w/ this method? I'm not sure I follow. This method will go away
when filterfilesystem supports nonrecursive create?
REVISION DETAIL
https://reviews.facebook.net/D1521
> support checksums in HBase block cache
> --------------------------------------
>
> Key: HBASE-5074
> URL: https://issues.apache.org/jira/browse/HBASE-5074
> Project: HBase
> Issue Type: Improvement
> Components: regionserver
> Reporter: dhruba borthakur
> Assignee: dhruba borthakur
> Attachments: D1521.1.patch, D1521.1.patch, D1521.2.patch,
> D1521.2.patch, D1521.3.patch, D1521.3.patch, D1521.4.patch, D1521.4.patch,
> D1521.5.patch, D1521.5.patch
>
>
> The current implementation of HDFS stores the data in one block file and the
> metadata(checksum) in another block file. This means that every read into the
> HBase block cache actually consumes two disk iops, one to the datafile and
> one to the checksum file. This is a major problem for scaling HBase, because
> HBase is usually bottlenecked on the number of random disk iops that the
> storage-hardware offers.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
