hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Wang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-6984) In Hadoop 3, make FileStatus serialize itself via protobuf
Date Tue, 06 Dec 2016 08:20:58 GMT

    [ https://issues.apache.org/jira/browse/HDFS-6984?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15724789#comment-15724789

Andrew Wang commented on HDFS-6984:

Thanks for revving this Chris. I don't have a ton of background here, but my review:

* Can we split the Serializable stuff into a separate change? Changing booleans into Boolean
objects is additional overhead, and combining these two changes makes it harder to review.
Also, I am not a Java serialization expert, but IIUC maintaining compatibility is difficult,
which means maybe people should just use PB anyway.
* The serialization loses the extra {{isEncrypted}} and {{FsPermission#getAclBit}} bits since
it calls {{toShort}} rather than {{toExtendedShort}}. Seems like we should save these, though
the fact we pack these bits into FsPermission is an internal implementation detail. Adding
new booleans to FileStatus might break cross-serialization with HdfsFileStatus though. What
is the usecase for cross-serialization?
* Could use some more extensive unit tests that test the default values and error conditions
with the shorts.
* Test nit: please also avoid the wildcard import.

Regarding Steve's comment on bounds checking, PB by default has a 64MB max message size. We
could use that as a reasonable upper-bound on the size of a FileStatus.

> In Hadoop 3, make FileStatus serialize itself via protobuf
> ----------------------------------------------------------
>                 Key: HDFS-6984
>                 URL: https://issues.apache.org/jira/browse/HDFS-6984
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>    Affects Versions: 3.0.0-alpha1
>            Reporter: Colin P. McCabe
>            Assignee: Colin P. McCabe
>              Labels: BB2015-05-TBR
>         Attachments: HDFS-6984.001.patch, HDFS-6984.002.patch, HDFS-6984.003.patch
> FileStatus was a Writable in Hadoop 2 and earlier.  Originally, we used this to serialize
it and send it over the wire.  But in Hadoop 2 and later, we have the protobuf {{HdfsFileStatusProto}}
which serves to serialize this information.  The protobuf form is preferable, since it allows
us to add new fields in a backwards-compatible way.  Another issue is that already a lot of
subclasses of FileStatus don't override the Writable methods of the superclass, breaking the
interface contract that read(status.write) should be equal to the original status.
> In Hadoop 3, we should just make FileStatus serialize itself via protobuf so that we
don't have to deal with these issues.  It's probably too late to do this in Hadoop 2, since
user code may be relying on the existing FileStatus serialization there.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org

View raw message