hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hadoop QA (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-4777) Write back to client 'incompatible' if we show up with wrong version
Date Sat, 12 Nov 2011 02:48:51 GMT

    [ https://issues.apache.org/jira/browse/HBASE-4777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13148959#comment-13148959

Hadoop QA commented on HBASE-4777:

-1 overall.  Here are the results of testing the latest attachment 
  against trunk revision .

    +1 @author.  The patch does not contain any @author tags.

    -1 tests included.  The patch doesn't appear to include any new or modified tests.
                        Please justify why no new tests are needed for this patch.
                        Also please list what manual steps were performed to verify this patch.

    -1 javadoc.  The javadoc tool appears to have generated -163 warning messages.

    +1 javac.  The applied patch does not increase the total number of javac compiler warnings.

    -1 findbugs.  The patch appears to introduce 50 new Findbugs (version 1.3.9) warnings.

    +1 release audit.  The applied patch does not increase the total number of release audit

    +1 core tests.  The patch passed unit tests in .

Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/236//testReport/
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/236//artifact/trunk/patchprocess/newPatchFindbugsWarnings.html
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/236//console

This message is automatically generated.
> Write back to client 'incompatible' if we show up with wrong version
> --------------------------------------------------------------------
>                 Key: HBASE-4777
>                 URL: https://issues.apache.org/jira/browse/HBASE-4777
>             Project: HBase
>          Issue Type: Bug
>            Reporter: stack
>            Assignee: stack
>         Attachments: 4777.txt
> We changed the RPC_VERSION to 4 in hbase-3939.  If a client comes in volunteering RPC_VERSION
is 3, currently, we'll log 'wrong version' but we'll close the connection; the client has
no chance of knowing why the server went away.
> Returning -1 as id up out of here is what causes the connection close:
> {code}
>     private void setupBadVersionResponse(int clientVersion) throws IOException {
>       String errMsg = "Server IPC version " + CURRENT_VERSION +
>       " cannot communicate with client version " + clientVersion;
>       ByteArrayOutputStream buffer = new ByteArrayOutputStream();
>       if (clientVersion >= 3) {
>         Call fakeCall =  new Call(-1, null, this, responder);
>         // Versions 3 and greater can interpret this exception
>         // response in the same manner
>         setupResponse(buffer, fakeCall, Status.FATAL,
>             null, VersionMismatch.class.getName(), errMsg);
>         responder.doRespond(fakeCall);
>       }
>     }
> {code}
> Instead, we need to return an id that does not close the connection so cilent gets chance
to read the response.
> Suggestion is that we return a 0 for the id.... the connection will stay up.
> If an old client and it sends the wrong version, it'll move on to do getProtocolVersion...
and will fail there.
> Other clients, e.g. asynchbase, if they get a response will have a response to switch
what they send to suit the new server.
> (There are other issues -- e.g. Invocation is versioned now -- but Benoit needs some
means of figuring whats on other side)

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


View raw message