[
http://issues.apache.org/jira/browse/HADOOP-538?page=comments#action_12438384 ]
Arun C Murthy commented on HADOOP-538:
--------------------------------------
Good points Doug; thanks!
Some thoughts...
I like the 'fallback' mechanism, it should ease hadoop adoption for people on
heterogenous environments, I'll get to work on this one.
However I'm a little skeptical about 'ship pre-built libhadoop.so' part for the
following reasons:
a) We will need to maintain versions for different h/w architectures and linux
distros (Definition of 'standard platforms'? -> cygwin and RHEL/ubuntu/FC{4|5}).
b) glibc compatibility (and also stuff like whether it is compiled with -nptl
or not and so on); agreed it isn't very common, but something which might bite
us once in a while. (Hmm... differing glibc across linux distros?)
c) Though we don't have it so far, a single line of C++ code in src/native will
make things far worse due to C++ ABI issues; means we will need to maintain
different versions of libhadoop for gcc-2x, gcc-3x, gcc-4x etc, even for
'standard platforms'.
Hence I'd like to propose we only have the 'fallback' mechanism - if you need
libhadoop.so, please build it yourself. We'll do all we can to ease the second
part.
I'll also start work to use the GNU build system (autoconf/automake/libtool)
for building libhadoop.so, this should ease the 'please build it yourself' part
for people.
Thoughts?
> Implement a nio's 'direct buffer' based wrapper over zlib to improve
> performance of java.util.zip.{De|In}flater as a 'custom codec'
> -----------------------------------------------------------------------------------------------------------------------------------
>
> Key: HADOOP-538
> URL: http://issues.apache.org/jira/browse/HADOOP-538
> Project: Hadoop
> Issue Type: Improvement
> Affects Versions: 0.6.1
> Reporter: Arun C Murthy
> Assigned To: Arun C Murthy
> Fix For: 0.7.0
>
> Attachments: HADOOP-538.patch, HADOOP-538_benchmarks.tgz
>
>
> There has been more than one instance where java.util.zip's {De|In}flater
> classes perform unreliably, a simple wrapper over zlib-1.2.3 (latest stable)
> using java.nio.ByteBuffer (i.e. direct buffers) should go a long way in
> alleviating these woes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
