ec2setup.txt instructions incomplete.

[rmichela]

Hello,
I'm trying to create an installation of the Data Science Toolkit on a fresh Ubuntu 12.04 installation running on VMWare in my datacenter. I've been using the steps listed in ec2setup.txt, but they seem to be missing a step.

Line 49 references the directory ~/sources/dstkdata which has not yet been created. In this directory I am supposed to untar gl_gpwfe_pdens_15_bil_25.tar.gz, but I have no idea where to get that file. glc2000_v1_1_Tiff.zip, SRTM_NE_250m.tif.zip, SRTM_W_250m.tif.zip, SRTM_SE_250m.tif.zip, tmean_30s_bil.zip also lack a source.

By chance, is there a version of the Data Science Toolkit that can run on VMWare? I understand you used to support this.

[petewarden]

(cc-ing dstk-users group)

Sorry about that Ryan! I'm in the middle of updating the build instructions
for some new data sources, so unfortunately they're in flux. Chris Child
has successfully imported the Vagrant box into VMWare though, he gives some
details in this thread:
https://groups.google.com/forum/?fromgroups=#!topic/dstk-users/0aC3Z4FfEiw
"All that needs to be done is to use Virtual Box to export the VM to .ova
format and just import it into VMware. That's it! It does assign it's own
IP though so users will have to login into the VM and find out what it is."
If you do still need to build your own from scratch, these should be the
stable instructions from the last release:
https://github.com/petewarden/dstk/blob/ee8f44df6ff395818c742f5523f583cf82884d3d/docs/ec2setup.txt

cheers,
Pete

On Tue, May 7, 2013 at 4:35 PM, Ryan Michela [email protected]:

Hello,
I'm trying to create an installation of the Data Science Toolkit on a
fresh Ubuntu 12.04 installation running on VMWare in my datacenter. I've
been using the steps listed in ec2setup.txt, but they seem to be missing a
step.

Line 49 references the directory ~/sources/dstkdata which has not yet been
created. In this directory I am supposed to untar
gl_gpwfe_pdens_15_bil_25.tar.gz, but I have no idea where to get that file.
glc2000_v1_1_Tiff.zip, SRTM_NE_250m.tif.zip, SRTM_W_250m.tif.zip,
SRTM_SE_250m.tif.zip, tmean_30s_bil.zip also lack a source.

By chance, is there a version of the Data Science Toolkit that can run on
VMWare? I understand you used to support this.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/17
.

[rmichela]

Thanks! I'm downloading the .box file now and let you know how it goes.

[rmichela]

Here are some other issues I noticed. Line 16 and 17 do not work from the ~ directory. I'm not sure how critical they are to the success of a DSTK server.

Also, I was given a 32 bit VM to work with. On line 13, /usr/lib/jvm/java-7-oracle/jre/lib/amd64/server needs to have i386 instead of amd64 for postgis to compile. The error message was extremely ambiguous.

[rmichela]

I'm skipping over the sentiment analysis data load for now and making progress. Here are some more notes I've collected. I'll put these in a PR after I've finished collecting them.

The passenger gem has been updated to 4.0.2
Line 102 should read: sudo bash -c 'echo "LoadModule passenger_module /var/lib/gems/1.8/gems/passenger-4.0.2/libout/apache2/mod_passenger.so" > /etc/apache2/mods-enabled/passenger.load'
Line 103 should read: sudo bash -c 'echo "PassengerRoot /var/lib/gems/1.8/gems/passenger-4.0.2" > /etc/apache2/mods-enabled/passenger.conf'

[rmichela]

Line 185: loadukpostcodes.sql has hard coded user path. Should have comment