hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jing Zhao (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-5281) COMMIT request should not block
Date Tue, 08 Oct 2013 17:58:46 GMT

    [ https://issues.apache.org/jira/browse/HDFS-5281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13789451#comment-13789451
] 

Jing Zhao commented on HDFS-5281:
---------------------------------

The patch looks pretty good. Some comments:

# In WriteManager.java,
{code}
+    if (openFileCtx != null) {
+      openFileCtx.checkCommit(dfsClient, commitOffset, channel, xid, preOpAttr);
     }
.....
+    Nfs3Utils.writeChannel(channel,
+        response.writeHeaderAndResponse(new XDR(), xid, new VerifierNone()),
+        xid);
{code}
Looks like we cannot always write the same response here. The response should depends on the
result of  OpenFileCtx#checkCommit?
# OpenFileCtx#checkCommit will also send response back to the client. Let's make the checkCommit
call only handles internal logic and let WriteManager#handleCommit to handle all the responses.
# Getter methods of CommitCtx need not be public.

> COMMIT request should not block
> -------------------------------
>
>                 Key: HDFS-5281
>                 URL: https://issues.apache.org/jira/browse/HDFS-5281
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: nfs
>            Reporter: Brandon Li
>            Assignee: Brandon Li
>         Attachments: HDFS-5281.001.patch, HDFS-5281.002.patch
>
>
> Currently Commit request is handled synchronously, blocked at most 30 seconds before
timeout. This JIRA is to make is asynchronous and thus it won't block other requests coming
from the same channel. 



--
This message was sent by Atlassian JIRA
(v6.1#6144)

Mime
View raw message