hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Aaron T. Myers (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-2802) Support for RW/RO snapshots in HDFS
Date Fri, 02 Nov 2012 17:29:16 GMT

    [ https://issues.apache.org/jira/browse/HDFS-2802?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13489564#comment-13489564

Aaron T. Myers commented on HDFS-2802:


bq. wrt to your alternate proposal, folks can get confused about which of the two series of
documents matched the patches in the Jira. Can you please rename yours as snapshot-alternate-design-proposal.{tex,pdf}
and also reflect that in the title inside the document.

Sure, I'll name the files differently from now on, though my hope is to get us collaborating
on a single design document, hence why I uploaded the .tex file as well. As Konst points out,
the two design proposals are currently very close to each other.


bq. Not sure how you will handle the snapshot deletion in the middle of the range. Say you
have snapshots 1 through 5, and then delete snapshot 3.

I agree that this should be discussed more in the document. The idea is that to delete a snapshot
from the perspective of the client, we'd only need to remove the info about that snapshot
from the given snapshottable directory. That could be done in constant time, and clients would
no longer be able to access that snapshot at that moment. Actual cleanup of files/directories
which are no longer referenced by any snapshot could then be done synchronously or asynchronously
at the NameNode.

bq. The idea of monotonously increasing snapshot ids seems to be the only major diversion
of your design from Suresh & Co. No doubt it should be discussed in details, but does
that justify/qualify-for submitting an alternative design.

I agree that the latest posted design doc is very close to the ones posted by Suresh and Nicholas,
but that was exactly the point - I was proposing a merged design that incorporated parts of
both. I would've just posted an updated doc which modifies the previously posted proposal,
as we often do with .patch files containing code, but that's difficult to do when only the
.pdf is uploaded, hence why I posted an entirely different document.

bq. You don't want to start Apache Hadoop design-document-wars here, do you?

Certainly not. I'm trying to collaborate on the design. The latest one I posted is an attempt
at that.
> Support for RW/RO snapshots in HDFS
> -----------------------------------
>                 Key: HDFS-2802
>                 URL: https://issues.apache.org/jira/browse/HDFS-2802
>             Project: Hadoop HDFS
>          Issue Type: New Feature
>          Components: data-node, name-node
>            Reporter: Hari Mankude
>            Assignee: Hari Mankude
>         Attachments: HDFS-2802.20121101.patch, HDFS-2802-meeting-minutes-121101.txt,
HDFSSnapshotsDesign.pdf, snap.patch, snapshot-design.pdf, snapshot-design.tex, snapshot-one-pager.pdf,
Snapshots20121018.pdf, Snapshots20121030.pdf
> Snapshots are point in time images of parts of the filesystem or the entire filesystem.
Snapshots can be a read-only or a read-write point in time copy of the filesystem. There are
several use cases for snapshots in HDFS. I will post a detailed write-up soon with with more

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

View raw message