Make error

[Alex]

Hi Danny,

I downloaded GraphLab 2.1 (featured version) and I'm trying to install it. I followed the steps in your blog but I am getting the following error when I type $make -j4

Please advise.

My system configuration is:

Hardware Overview:

  Model Identifier: MacBookPro10,1

  Processor Name: Intel Core i7

  Processor Speed: 2.6 GHz

  Number of Processors: 1

  Total Number of Cores: 4

  Memory: 16 GB

Software  OS X 10.8.3 (12D78)

ERROR:

$ Scanning dependencies of target hadoop

Scanning dependencies of target libevent

Scanning dependencies of target libbz2

Scanning dependencies of target eigen

[  0%] [  0%] [  2%] [  2%] Creating directories for 'libbz2'

Creating directories for 'libevent'

Creating directories for 'hadoop'

Creating directories for 'eigen'

[  2%] [  5%] [  5%] [  5%] Performing download step (download, verify and extract) for 'eigen'

Performing download step (download, verify and extract) for 'libevent'

Performing download step (download, verify and extract) for 'hadoop'

Performing download step (download, verify and extract) for 'libbz2'

-- downloading...

     src='http://www.bzip.org/1.0.6/bzip2-1.0.6.tar.gz'

     dst='/Users/a1021997/Documents/c/graphlabapi/deps/libbz2/src/bzip2-1.0.6.tar.gz'

     timeout='none'

-- downloading...

     src='http://iweb.dl.sourceforge.net/project/levent/libevent/libevent-2.0/libevent-2.0.18-stable.tar.gz'

     dst='/Users/a1021997/Documents/c/graphlabapi/deps/event/src/libevent-2.0.18-stable.tar.gz'

     timeout='none'

-- downloading...

     src='http://graphlab.org/deps/eigen_3.1.2.tar.bz2'

     dst='/Users/a1021997/Documents/c/graphlabapi/deps/eigen/src/eigen_3.1.2.tar.bz2'

     timeout='none'

-- downloading...

     src='http://graphlab.org/deps/hadoop-1.0.1.tar.gz'

     dst='/Users/a1021997/Documents/c/graphlabapi/deps/hadoop/src/hadoop-1.0.1.tar.gz'

     timeout='none'

CMake Error at eigen-stamp/download-eigen.cmake:9 (file):

  file DOWNLOAD HASH mismatch

    for file: [/Users/a1021997/Documents/c/graphlabapi/deps/eigen/src/eigen_3.1.2.tar.bz2]

      expected hash: [e9c081360dde5e7dcb8eba3c8430fde2]

        actual hash: [d41d8cd98f00b204e9800998ecf8427e]

CMake Error at libevent-stamp/download-libevent.cmake:9 (file):

  file DOWNLOAD HASH mismatch

    for file: [/Users/a1021997/Documents/c/graphlabapi/deps/event/src/libevent-2.0.18-stable.tar.gz]

      expected hash: [aa1ce9bc0dee7b8084f6855765f2c86a]

        actual hash: [d41d8cd98f00b204e9800998ecf8427e]

CMake Error at libbz2-stamp/download-libbz2.cmake:9 (file):

  file DOWNLOAD HASH mismatch

    for file: [/Users/a1021997/Documents/c/graphlabapi/deps/libbz2/src/bzip2-1.0.6.tar.gz]

      expected hash: [00b516f4704d4a7cb50a1d97e6e8e15b]

        actual hash: [d41d8cd98f00b204e9800998ecf8427e]

CMake Error at hadoop-stamp/download-hadoop.cmake:9 (file):

  file DOWNLOAD HASH mismatch

    for file: [/Users/a1021997/Documents/c/graphlabapi/deps/hadoop/src/hadoop-1.0.1.tar.gz]

      expected hash: [e627d9b688c4de03cba8313bd0bba148]

        actual hash: [d41d8cd98f00b204e9800998ecf8427e]

make[2]: *** [../deps/event/src/libevent-stamp/libevent-download] Error 1

make[2]: *** [../deps/hadoop/src/hadoop-stamp/hadoop-download] Error 1make[2]: 

*** [../deps/eigen/src/eigen-stamp/eigen-download] Error 1

make[2]: *** [../deps/libbz2/src/libbz2-stamp/libbz2-download] Error 1

make[1]: *** [CMakeFiles/libevent.dir/all] Error 2

make[1]: *** Waiting for unfinished jobs....

make[1]: *** [CMakeFiles/eigen.dir/all] Error 2

make[1]: *** [CMakeFiles/libbz2.dir/all] Error 2

make[1]: *** [CMakeFiles/hadoop.dir/all] Error 2

make: *** [all] Error 2

[Alex]

Hi Danny,

Any help on how to resolve this error will be appreciated.

Thanks,

Alex

[Yucheng Low]

Hi,

Notice that the hash of both files are the same: d41d8cd98f00b204e9800998ecf8427e
This is  the hash of the empty string.

You are probably behind a proxy of some sort, and CMake is unable to download the dependencies.

You probably have a misconfigured proxy issue.

Does wget work correctly?

(Try running wget http://graphlab.org/deps/libjson_7.6.0.zip and see if it downloads correctly)

Yucheng

On Mar 28, 2013, at 8:51 AM, "Berry, Jonathan" <jbe...@sandia.gov> wrote:

Yucheng,

Can you make sense of this?  The files were downloaded by the build process.

-Jon

CMake Error at eigen-stamp/download-eigen.cmake:9 (file):
file DOWNLOAD HASH mismatch

  for file: [/home/jberry/graphlabapi/deps/eigen/src/eigen_3.1.2.tar.bz2]

    expected hash: [e9c081360dde5e7dcb8eba3c8430fde2]
      actual hash: [d41d8cd98f00b204e9800998ecf8427e]

CMake Error at libjson-stamp/download-libjson.cmake:9 (file):
file DOWNLOAD HASH mismatch

  for file: [/home/jberry/graphlabapi/deps/json/src/libjson_7.6.0.zip]
    expected hash: [dcb326038bd9b710b8f717580c647833]
      actual hash: [d41d8cd98f00b204e9800998ecf8427e]

________________________________________
From: Yucheng Low [yl...@cs.cmu.edu]
Sent: Wednesday, March 27, 2013 12:34 PM
To: Berry, Jonathan
Cc: Kolda, Tamara G
Subject: Re: [EXTERNAL] question from last week

Hi,

That is odd. We have not had complaints about difficulty downloading stuff in a very long time.
In fact, I just did a successful reclone and rebuild of everything.

Are you having internet connectivity issues?
The hashes on our server match exactly correctly too.
Are you pulling across a proxy? I have had reports before that on certain proxy setups, downloads will obtain some sort of proxy error page instead of the actual contents.
That might also affect the mercurial clone process.

Yucheng

On Mar 27, 2013, at 11:10 AM, "Berry, Jonathan" <jbe...@sandia.gov> wrote:

Yucheng,

I'm trying to make a sincere effort to get graphlabapi built, but I'm not having much success.

I gave up on pulling from the repository (always hangs) and got the tarball.  I'm now trying to build, and I advanced past a couple of cmake issues and finally got configured.  However, the first thing that make does is try to pull a lot of dependencies from the web.  These typically either time out (e.g. sourceforge) or fail with mismatched hashes (e.g. libjson_7.6.0.zip, eigen_3.1.2.tar.bz2 from your site).  Are you sure that the software is ready to get?

Regards,
Jon Berry

-----Original Message-----
From: Yucheng Low [mailto:ylow@cs.cmu.edu]
Sent: Monday, March 25, 2013 2:14 PM
To: Berry, Jonathan
Subject: Re: [EXTERNAL] question from last week

No... But sometimes googlecode acts up a bit.
Continuous hammering works.

Yucheng

On Mar 25, 2013, at 12:52 PM, "Berry, Jonathan" <jbe...@sandia.gov> wrote:

Yucheng,

I tried to get graphlab just now (first getting mercurial), but the clone line just hangs.  Is that a common problem?

Thanks,
-Jon

-----Original Message-----
From: Yucheng Low [mailto:ylow@cs.cmu.edu]
Sent: Tuesday, March 05, 2013 11:47 AM
To: Berry, Jonathan; Pinar, Ali; Kolda, Tamara G
Cc: Joseph Gonzalez
Subject: [EXTERNAL] Re: question from last week

Hi,

By your definition of counting versus enumeration, we are definitely doing enumeration.
The overall procedure is quite simple to describe in the undirected case.
1: On gather/apply: each vertex collects a list of all its neighbors
2: On scatter, for each edge, we compute the intersection between the lists on both ends of the edge, and record the size of the intersection.

The trick to make the triangle counting implementation is to keep memory requirements low, (the std::set or boost::unordered_set is kind of heavy), and to compute the intersection quickly.
The implementation uses a cuckoo-set when the number of list elements are large, and a simple sorted array when the number of list elements are small.

Yucheng

On Mar 5, 2013, at 9:30 AM, Joseph Gonzalez <jego...@cs.cmu.edu> wrote:

Hi Berry,

The 90 seconds is for counting the number of triangles of each of the
4 kinds that pass through each vertex for the twitter graph (41M
vertices, 1.4B edges).  We then return for each vertex the count of
each of the 4 motifs.  This is done using the code here:

http://code.google.com/p/graphlabapi/source/browse/toolkits/graph_ana
l
ytics/directed_triangle_count.cpp

A more detailed description is provided here:

http://docs.graphlab.org/graph_analytics.html#graph_analytics_triangl
e
_directed

We are actually not doing anything really clever and in both cases we
simply enumerate all the possible triangles.

I have cc'ed Yucheng who wrote the code for triangle counting and ran
those experiments so he can fill you in on the details.  Assuming you
can get GraphLab running on your system (requires C++/MPI) it might
be worth trying since in addition to triangle counting we have a
bunch of other graph analytics packages.

http://docs.graphlab.org/graph_analytics.html

Let us know if you have any questions since it would be fun to collaborate.

Joey



On Mon, Mar 4, 2013 at 3:48 PM, Berry, Jonathan <jbe...@sandia.gov> wrote:

Greetings Joseph,



My colleagues Tammy Kolda and Ali Pinar spoke with you at the SIAM
conference last week and heard of your GraphLab computation of triangles in
twitter-2010 in 90 seconds on some small number of processors.   That
certainly sounds like an interesting computation that I might need
to learn about.  To start, was this enumerating triangles or
counting them?  By enumerating, I mean that the computation sees
each triangle, whereas counting could be done with eigenvalues, etc.
I've got a computation that enumerates all triangles and takes a
long time on the twitter instance (34B
triangles) even in shared memory.  For each of those 34B triangles
it updates a set of vertex and edge properties.  If I could do the
same thing in GraphLab quickly that would be quite interesting.



Regards,

Jon Berry

Sandia Labs.

--
You received this message because you are subscribed to the Google Groups "GraphLab API" group.
To unsubscribe from this group and stop receiving emails from it, send an email to graphlabapi...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.