Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Thu, 21 Dec 2017 22:00:00 +0000 (UTC)
From: "stack (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.13126655.1513892256000.514891.1513893600053@Atlassian.JIRA>
In-Reply-To: <JIRA.13126655.1513892256000@Atlassian.JIRA>
References: <JIRA.13126655.1513892256000@Atlassian.JIRA> <JIRA.13126655.1513892256107@jira-lw-us.apache.org>
Subject: [jira] [Commented] (HBASE-19585) [WAL] "Unhandled: Bad type on
 operand stack" PBHelperClient.convert failing on
 HdfsProtos$ContentSummaryProto
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Thu, 21 Dec 2017 22:00:10 -0000


    [ https://issues.apache.org/jira/browse/HBASE-19585?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16300593#comment-16300593 ] 

stack commented on HBASE-19585:
-------------------------------

Happens on jdk8_151 too. Happens even if I clean out all data and start fresh; i.e. not on a 0.98 written-hbase. Line #414 is the forName in the below... But looks like we are on to the convert below. This seemed to work previously when I ran w/ asyncfswal enabled..

{code}
  private static PBHelper createPBHelper() throws NoSuchMethodException {
    Class<?> helperClass;
    String clazzName = "org.apache.hadoop.hdfs.protocolPB.PBHelperClient";
    try {
      helperClass = Class.forName(clazzName);
    } catch (ClassNotFoundException e) {
      helperClass = org.apache.hadoop.hdfs.protocolPB.PBHelper.class;
      LOG.debug("" + clazzName + " not found (Hadoop is pre-2.8.0?); using " +
          helperClass.toString() + " instead.");
    }
    Method convertEBMethod = helperClass.getMethod("convert", ExtendedBlock.class);
    Method convertTokenMethod = helperClass.getMethod("convert", Token.class);
    return new PBHelper() {

      @Override
      public ExtendedBlockProto convert(ExtendedBlock b) {
        try {
          return (ExtendedBlockProto) convertEBMethod.invoke(null, b);
        } catch (IllegalAccessException | InvocationTargetException e) {
          throw new RuntimeException(e);
        }
      }

      @Override
      public TokenProto convert(Token<?> tok) {
        try {
          return (TokenProto) convertTokenMethod.invoke(null, tok);
        } catch (IllegalAccessException | InvocationTargetException e) {
          throw new RuntimeException(e);
        }
      }
    };
  }
{code}


> [WAL] "Unhandled: Bad type on operand stack" PBHelperClient.convert failing on HdfsProtos$ContentSummaryProto
> -------------------------------------------------------------------------------------------------------------
>
>                 Key: HBASE-19585
>                 URL: https://issues.apache.org/jira/browse/HBASE-19585
>             Project: HBase
>          Issue Type: Bug
>            Reporter: stack
>
> Testing, RS crashes soon after startup with below cryptic mess. This is the branch-2 started over a 0.98 data. [~Apache9] You have a clue sir?
> {code}
> 595126 2017-12-21 13:09:38,058 INFO  [regionserver/ve0528.halxg.cloudera.com/10.17.240.22:16020] wal.AbstractFSWAL: WAL configuration: blocksize=128 MB, rollsize=121.60 MB, prefi       x=ve0528.halxg.cloudera.com%2C16020%2C1513890565537, suffix=, logDir=hdfs://ve0524.halxg.cloudera.com:8020/hbase/WALs/ve0528.halxg.cloudera.com,16020,1513890565537, archiv       eDir=hdfs://ve0524.halxg.cloudera.com:8020/hbase/oldWALs
> 595127 2017-12-21 13:09:38,107 ERROR [regionserver/ve0528.halxg.cloudera.com/10.17.240.22:16020] regionserver.HRegionServer: ***** ABORTING region server ve0528.halxg.cloudera.co       m,16020,1513890565537: Unhandled: Bad type on operand stack
> 595128 Exception Details:
> 595129   Location:
> 595130     org/apache/hadoop/hdfs/protocolPB/PBHelperClient.convert(Lorg/apache/hadoop/hdfs/protocol/proto/HdfsProtos$ContentSummaryProto;)Lorg/apache/hadoop/fs/ContentSummary; @       98: invokestatic
> 595131   Reason:
> 595132     Type 'org/apache/hadoop/fs/ContentSummary$Builder' (current frame, stack[1]) is not assignable to 'org/apache/hadoop/fs/QuotaUsage$Builder'
> 595133   Current Frame:
> 595134     bci: @98
> 595135     flags: { }
> 595136     locals: { 'org/apache/hadoop/hdfs/protocol/proto/HdfsProtos$ContentSummaryProto', 'org/apache/hadoop/fs/ContentSummary$Builder' }
> 595137     stack: { 'org/apache/hadoop/hdfs/protocol/proto/HdfsProtos$StorageTypeQuotaInfosProto', 'org/apache/hadoop/fs/ContentSummary$Builder' }
> 595138   Bytecode:
> 595139     0x0000000: 2ac7 0005 01b0 bb03 3159 b703 324c 2b2a
> 595140     0x0000010: b603 33b6 0334 2ab6 0335 b603 362a b603
> 595141     0x0000020: 37b6 0338 2ab6 0339 b603 3a2a b603 3bb6
> 595142     0x0000030: 033c 2ab6 033d b603 3e2a b603 3fb6 0340
> 595143     0x0000040: 2ab6 0341 b603 422a b603 43b6 0344 2ab6
> 595144     0x0000050: 0345 b603 4657 2ab6 0347 9900 0b2a b603
> 595145     0x0000060: 482b b803 492b b603 4ab0
> 595146   Stackmap Table:
> 595147     same_frame(@6)
> 595148     append_frame(@101,Object[#2126])
> 595149  *****
> {code}
> 2.8.2 hadoop.


--
This message was sent by Atlassian JIRA
(v6.4.14#64029)