Greenplum DB files

[Xavier Puig Fernandez]

Hello,

I am trying to set Chorus for development but I can't find where are the Greenplum DB and Greenplum Hadoop files that are specified in the Get Started Guide. All the links that I've found seem to be broken.

Could you provide a website to download the files?

[Robbie GIll]

We do not distribute the those files, but you can find current versions on the Pivotal website (https://network.gopivotal.com/products/pivotal-gpdb).

The Virtualization section is primarily used for running unit and integration tests during development, but if you have access to other gpdb/hadoop sources, those can be used instead. 

See also: https://groups.google.com/d/msg/chorus-dev/V-TF0OtKRlI/fyCUgouhMC0J

[Xavier Puig Fernandez]

Thank you so much.

I have installed vagrant and virtualbox and have run vagrant up into the chorus_appliance directory. However, when I run "rake development:init" I get the following error:

The driver encountered an unknown error: org.postgresql.util.PSQLException: Connection refused. Check that the hostname and port are correct and that the postmaster is accepting TCP/IP connections.

 

The problem is that the guide says that such command should initialize a new database in postgres-db but this directory doesn't exist at this point of the installation.

Do you have any idea about what could be the problem?

El dimecres 2 de juliol de 2014 16:27:45 UTC+2, Xavier Puig Fernandez va escriure:

[Robbie Gill]

It sounds like something failed during the postgres initialization. That step should create the postgres-db directory. If the postgres-db directory exists after the failure, try looking at the server.log file located within - it could provide details as to why postgres initialization failed. 

To retry, remove the postgres-db directory entirely before running rake development:init again.  

If you are are a mac, I have seen postgres initialization fail if you do not increase the shared memory settings as described in getting started. Also, you can run rake with the --trace flag to get more verbose output.

Hope that helps.