hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-1975) HA: Support for sharing the namenode state from active to standby.
Date Thu, 20 Oct 2011 20:19:10 GMT

    [ https://issues.apache.org/jira/browse/HDFS-1975?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13131983#comment-13131983
] 

Todd Lipcon commented on HDFS-1975:
-----------------------------------

- can we move EditLogTailer to the ha package?
- should probably have some brief javadoc there. Also, if it is a public class, it needs InterfaceAudience/Stability
annotations
- why do we sleep 60 seconds between tails? I think keeping the standby as up to date as possible
is important. Though it's not an immediate goal of today's project, we should keep in mind
the secondary goal of serving stale reads from the standby.
- using the terminology "sharedEditDirs" implies that we only support directories here. Instead,
shouldn't we call it "sharedEditUris"? Same with the configs, etc.
- the code in {{stopReadingEditLogs}} seems really race prone. We need better inter-thread
coordination than just sleeping.
- The name {{waitForGenStamp}} implies that it waits for something, but in fact this is just
{{isGenStampInFuture}}
- need license on PendingDataNodeMessages
- need javadoc in lots of spots - explain the purposes of the new class, etc
- getMaxGsInBlockList could be moved to a static method in BlockListAsLongs

- DataNodeMessage and subclasses: make the fields final

- needs unit tests - there are some in my earlier patch for this issue that could be used
to verify EditLogTailer, I think.

I want to also do some thinking on synchronization for the datanode messages, etc. Will write
more later.

                
> HA: Support for sharing the namenode state from active to standby.
> ------------------------------------------------------------------
>
>                 Key: HDFS-1975
>                 URL: https://issues.apache.org/jira/browse/HDFS-1975
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: name-node
>            Reporter: Suresh Srinivas
>            Assignee: Jitendra Nath Pandey
>         Attachments: HDFS-1975-HA.2.patch, HDFS-1975-HA.patch, hdfs-1975.txt, hdfs-1975.txt
>
>
> To enable hot standby namenode, the standby node must have current information for -
namenode state (image + edits) and block location information. This jira addresses keeping
the namenode state current in the standby node. To do this, the proposed solution in this
jira is to use a shared storage to store the namenode state. 
> Note one could also build an alternative solution by augmenting the backup node. A seperate
jira could explore this.

--
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

        

Mime
View raw message