[ https://issues.apache.org/jira/browse/HDFS-4710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13640942#comment-13640942 ]
Gopal V commented on HDFS-4710: ------------------------------- I did consider that as a fix, but a regular BufferedInputStream isn't seekable (particularly backwards), which would need to be reimplemented on top of it. This needs some of the same ByteBuffer juggling used in the verify checksum mode - to buffer data in chunks, but drop them when seeking. > Turning off HDFS short-circuit checksums unexpectedly slows down Hive > --------------------------------------------------------------------- > > Key: HDFS-4710 > URL: https://issues.apache.org/jira/browse/HDFS-4710 > Project: Hadoop HDFS > Issue Type: Bug > Components: hdfs-client > Affects Versions: 2.0.4-alpha > Environment: Centos (EC2) + short-circuit reads on > Reporter: Gopal V > Priority: Minor > Labels: perfomance > > When short-circuit reads are on, HDFS client slows down when checksums are > turned off. > With checksums on, the query takes 45.341 seconds and with it turned off, it > takes 56.345 seconds. This is slower than the speeds observed when > short-circuiting is turned off. > The issue seems to be that FSDataInputStream.readByte() calls are directly > transferred to the disk fd when the checksums are turned off. > Even though all the columns are integers, the data being read will be read > via DataInputStream which does > {code} > public final int readInt() throws IOException { > int ch1 = in.read(); > int ch2 = in.read(); > int ch3 = in.read(); > int ch4 = in.read(); > {code} > To confirm, an strace of the Yarn container shows > {code} > 26690 read(154, "B", 1) = 1 > 26690 read(154, "\250", 1) = 1 > 26690 read(154, ".", 1) = 1 > 26690 read(154, "\24", 1) = 1 > {code} > To emulate this without the entirety of Hive code, I have written a simpler > test app > https://github.com/t3rmin4t0r/shortcircuit-reader > The jar will read a file in -bs <n> sized buffers. Running it with 1 byte > blocks gives similar results to the Hive test run. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira