hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Raghu Angadi <rang...@apache.org>
Subject Re: Problem reading HDFS block size > 1GB
Date Thu, 01 Oct 2009 22:48:13 GMT
Vinay,

This issue came up before (
http://www.mail-archive.com/core-dev@hadoop.apache.org/msg35620.html ) . I
think we should fix this soon. Dhruba filed a jira (
https://issues.apache.org/jira/browse/HDFS-96 ) .  Not all errors reported
here fixed by the patch attached there. Could we discuss this on the jira.
For now you could try the path there in addition to your fix.

I think supporting large block size requires a fixes like the one for
FSInputChecker below, BlockSender fix in HDFS-96. I think it would not
require any major structural changes. Please attach you findings to the
jira. I can help.

Regd read timeout from DN, could you check log on the datanode as well?

Raghu.

On Tue, Sep 29, 2009 at 10:52 AM, Vinay Setty <vinay.setty@gmail.com> wrote:

> Hi all,
> We are running Yahoo distribution of Hadoop based on Hadoop 0.20.0-2787265
> .
> On a 10 nodes cluster with OpenSUSE Linux Operating System. We have HDFS
> configured with Block Size 5GB (This is for our experiments). But we are
> facing following problems when we try reading the data beyond 1GB from 5GB
> input split.
>
> *1) Problem in DFSClient*
>  When we read the text data, the map reduce job was getting stuck after
> reading about first 1GB of data from the split. Thereafter it was unable to
> read anymore data using IOUtils.readFully() and after 10 such minutes
> Hadoop
> got timed out. Then we tried to manually skip first 2GB from the 5GB split
> and start reading from 2GB offset inside the split. There skipping was ok
> but reading the first line after skipping 2GB gave the following exception:
>
> java.lang.IndexOutOfBoundsException
> at org.apache.hadoop.fs.FSInputChecker.read(FSInputChecker.java:151)
> at org.apache.hadoop.hdfs.DFSClient$BlockReader.read(DFSClient.java:1118)
>  at
>
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.readBuffer(DFSClient.java:1666)
> at
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.read(DFSClient.java:1716)
>  at java.io.DataInputStream.read(DataInputStream.java:132)
> at org.apache.hadoop.io.IOUtils.readFully(IOUtils.java:100)
>  at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderRecordReader.<init>(LoaderRecordReader.java:101)
> at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderInputFormat.getRecordReader(LoaderInputFormat.java:45)
>  at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:337)
> at org.apache.hadoop.mapred.MapTask.run(MapTask.java:306)
>  at org.apache.hadoop.mapred.Child.main(Child.java:170)
>
> This was particularly strange. We looked at line number 151 of
> FSInputChecker where the exception was thrown and it looks like below:
>
> public synchronized int read(byte[] b, int off, int len) throws IOException
> {
>    if ((off | len | (off + len) | (b.length - (off + len))) < 0) {
>      throw new IndexOutOfBoundsException();
>    }
>
> It looks like parameters "off" or "len" got negative. We traced the stack
> back and found following interesting thing at lines 1715 and 1716 in
> DFSClient:
>
>            int realLen = Math.min(len, (int) (blockEnd - pos + 1));
>            int result = readBuffer(buf, off, realLen);
>
> The first line takes int of (blockEnd - pos +1) where blockEnd is absolute
> end position of the block (split) in the file and pos is the current
> position, both as long. But after taking their integer value it might
> overflow! and make realLen negative, thereby giving
> IndexOutOfBoudsException.
>
> This problem we fixed by modifying  DFSClient.java:1715 to
>
> int realLen = (int)Math.min(len,  (blockEnd - pos + 1));
>
> *2) Problem with  IOUtils.skipFully() *
>
> We fixed the problem of casting from long to int, but we got a new problem,
> i.e. we are not able to read more than 1GB of data. When we
> did  IOUtils.skipFully() and skipped 1GB and treid to read from there, we
> following exception.
>
> java.io.EOFException
>        at java.io.DataInputStream.readInt(DataInputStream.java:375)
>        at
> org.apache.hadoop.hdfs.DFSClient$BlockReader.readChunk(DFSClient.java:1218)
>        at
>
> org.apache.hadoop.fs.FSInputChecker.readChecksumChunk(FSInputChecker.java:237)
>        at org.apache.hadoop.fs.FSInputChecker.fill(FSInputChecker.java:176)
>        at
> org.apache.hadoop.fs.FSInputChecker.read1(FSInputChecker.java:193)
>        at org.apache.hadoop.fs.FSInputChecker.read(FSInputChecker.java:158)
>        at
> org.apache.hadoop.hdfs.DFSClient$BlockReader.read(DFSClient.java:1117)
>        at
>
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.readBuffer(DFSClient.java:1665)
>        at
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.read(DFSClient.java:1720)
>        at java.io.DataInputStream.read(DataInputStream.java:132)
>        at org.apache.hadoop.io.IOUtils.readFully(IOUtils.java:100)
>        at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderRecordReader.<init>(LoaderRecordReader.java:107)
>        at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderInputFormat.getRecordReader(LoaderInputFormat.java:45)
>        at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:336)
>        at org.apache.hadoop.mapred.MapTask.run(MapTask.java:305)
>        at org.apache.hadoop.mapred.Child.main(Child.java:170)
>
> 3) *java.net.SocketTimeoutException*
>
> Since there seemed to be some problem with IOUtils.skipFully() we started
> seeking just before reading each record. We do not get End of File
> exception
> anymore but get the following new exception after reading approx 1810517942
> bytes (around 1.5GB).
>
> java.net.SocketTimeoutException: 60000 millis timeout while waiting for
> channel to be ready for read. ch :
> java.nio.channels.SocketChannel[connected
> local=/134.96.223.140:48255remote=/134.96.223.140:50010]
>  at
>
> org.apache.hadoop.net.SocketIOWithTimeout.doIO(SocketIOWithTimeout.java:164)
> at org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:155)
>  at
> org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:128)
> at java.io.BufferedInputStream.fill(BufferedInputStream.java:218)
>  at java.io.BufferedInputStream.read(BufferedInputStream.java:237)
> at java.io.DataInputStream.readInt(DataInputStream.java:370)
>  at
> org.apache.hadoop.hdfs.DFSClient$BlockReader.readChunk(DFSClient.java:1218)
> at
>
> org.apache.hadoop.fs.FSInputChecker.readChecksumChunk(FSInputChecker.java:237)
>  at org.apache.hadoop.fs.FSInputChecker.fill(FSInputChecker.java:176)
> at org.apache.hadoop.fs.FSInputChecker.read1(FSInputChecker.java:193)
>  at org.apache.hadoop.fs.FSInputChecker.read(FSInputChecker.java:158)
> at org.apache.hadoop.hdfs.DFSClient$BlockReader.read(DFSClient.java:1117)
>  at
>
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.readBuffer(DFSClient.java:1665)
> at
> org.apache.hadoop.hdfs.DFSClient$DFSInputStream.read(DFSClient.java:1720)
>  at java.io.DataInputStream.read(DataInputStream.java:132)
> at org.apache.hadoop.io.IOUtils.readFully(IOUtils.java:100)
>  at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderRecordReader.next(LoaderRecordReader.java:163)
> at
>
> edu.unisb.cs.mapreduce.binary.load.LoaderRecordReader.next(LoaderRecordReader.java:1)
>  at
>
> org.apache.hadoop.mapred.MapTask$TrackedRecordReader.moveToNext(MapTask.java:191)
> at
> org.apache.hadoop.mapred.MapTask$TrackedRecordReader.next(MapTask.java:175)
>  at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:48)
> at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:356)
>  at org.apache.hadoop.mapred.MapTask.run(MapTask.java:305)
> at org.apache.hadoop.mapred.Child.main(Child.java:170)
>
> Steps to reproduce:
>
>   1. Configure HDFS with 5GB block size
>   2. Load text data of > 5GB to HDFS
>   3. Run Grep or wordcount on the data
>
> We have following questions:
>
>
>   1. Does any body know what is the maximum block size and split size
>   supported by HDFS?
>   2. Did any of you guys try something similar?
>   3. Did any of you guys get similar problems? If so can you please tell us
>   how you resolved it?
>
> Thank you,
> Vinay Setty
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message