hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Zhe Zhang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-9079) Erasure coding: preallocate multiple generation stamps and serialize updates from data streamers
Date Fri, 23 Oct 2015 17:27:27 GMT

    [ https://issues.apache.org/jira/browse/HDFS-9079?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14971412#comment-14971412
] 

Zhe Zhang commented on HDFS-9079:
---------------------------------

Thanks Nicholas for the comment.
bq. > ... => 2) Asks NN for new GS => 3) Gets new GS from NN => ...
bq. What is the difference between #2 and #3? Is it just a single RPC?
Yes it's a single RPC. I listed them as 2 steps because other events could happen between
#2 and #3. E.g. while {{streamer_i}} is waiting for NN response {{streamer_j}} might start
step #2.

bq. Do you mean that client may update GS without letting NN knowing it?
More details of the proposed protocol can be found [here | https://issues.apache.org/jira/browse/HDFS-9040?focusedCommentId=14741972&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14741972].
When {{streamer_i}} encounters a DN failure, the coordinator will ask other streamers to update
their DN to increment GS. After all healthy DNs acknowledge that they have bumped their local
GSes, the coordinator send {{updatePipeline}} RPC to NN to update the NN's copy of GS. So
there will never be "false stale" -- fresh replica being considered as stale. 

bq. How to save step #1?
Good catch, I meant saving steps 2~3.

> Erasure coding: preallocate multiple generation stamps and serialize updates from data
streamers
> ------------------------------------------------------------------------------------------------
>
>                 Key: HDFS-9079
>                 URL: https://issues.apache.org/jira/browse/HDFS-9079
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: erasure-coding
>    Affects Versions: HDFS-7285
>            Reporter: Zhe Zhang
>            Assignee: Zhe Zhang
>         Attachments: HDFS-9079-HDFS-7285.00.patch, HDFS-9079.01.patch, HDFS-9079.02.patch,
HDFS-9079.03.patch, HDFS-9079.04.patch, HDFS-9079.05.patch
>
>
> A non-striped DataStreamer goes through the following steps in error handling:
> {code}
> 1) Finds error => 2) Asks NN for new GS => 3) Gets new GS from NN => 4) Applies
new GS to DN (createBlockOutputStream) => 5) Ack from DN => 6) Updates block on NN
> {code}
> To simplify the above we can preallocate GS when NN creates a new striped block group
({{FSN#createNewBlock}}). For each new striped block group we can reserve {{NUM_PARITY_BLOCKS}}
GS's. Then steps 1~3 in the above sequence can be saved. If more than {{NUM_PARITY_BLOCKS}}
errors have happened we shouldn't try to further recover anyway.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message