Alejandro Cabrera f6c321dd8b feat: integrate shard storage with transport
This patch refines the interface and storage implementation defined in
the last patch and integrates it with the transport layer.

A few updates have been made:
- 'name' -> 'href' for listing shards
- limiting, markers, and detailed are all used
- use of common_utils.fields to clean up shards transport PATCH
- add missing init for schemas
- fix schema issues found: 'location' -> 'uri', __init__.py
- shard resource correctly implements PUT semantics (replaces)

Transport: the admin API concept has been expanded to include
functionality from the public interface *in addition* to admin
functionality. Part of the rationale behind this is to simplify unit
testing. The other part of this is that an admin should be able to do
everything a normal user can do in addition to their special
functions.

Storage: now divided into control and data plane. The bootstrap passes
a control driver down to the transport which *can* be used for
endpoints as needed.

A test suite has been added that exercises the functionality from the
transport side of the shard registry resource.

Finally, the way the FaultyStorage driver tests were handled was
changed. Something about the setattr magic in that suite's setup made
it such that *all* tests would use the Faulty storage driver. This is
possibly related to the use of lazy_property decorators. To address
this issue, this patch promotes the faulty storage driver to setup.cfg
visibility and removes the setattrs.

Change-Id: I5b8cdb3a11d29422762b52f1e15e33167eecb867
Partitally-implements: blueprint storage-sharding
Partially-Closes: 1241686
Closes-Bug: 1243898
2013-10-31 09:12:54 -04:00
2013-08-14 16:10:08 -05:00
2012-11-01 09:52:20 +01:00

Marconi

Message queuing service for OpenStack

Running a local Marconi server with MongoDB

Note: These instructions are for running a local instance of Marconi and not all of these steps are required. It is assumed you have MongoDB installed and running.

  1. From your home folder create the ~/.marconi folder and clone the repo:

    $ cd
    $ mkdir .marconi
    $ git clone https://github.com/openstack/marconi.git
  2. Copy the Marconi config files to the directory ~/.marconi:

    $ cp marconi/etc/marconi-proxy.conf-sample ~/.marconi/marconi-proxy.conf
    $ cp marconi/etc/marconi-queues.conf-sample ~/.marconi/marconi-queues.conf
    $ cp marconi/etc/logging.conf-sample ~/.marconi/logging.conf
  3. Find the [drivers:storage:mongodb] section in ~/.marconi/marconi-queues.conf and modify the URI to point to your local mongod instance:

    uri = mongodb://$MONGODB_HOST:$MONGODB_PORT
  4. For logging, find the [DEFAULT] section in ~/.marconi/marconi-queues.conf and modify as desired:

    log_file = server.log
  5. Change directories back to your local copy of the repo:

    $ cd marconi
  6. Run the following so you can see the results of any changes you make to the code without having to reinstall the package each time:

    $ pip install -e .
  7. Start the Marconi server:

    $ marconi-server
  8. Test out that Marconi is working by creating a queue:

    $ curl -i -X PUT http://127.0.0.1:8888/v1/queues/samplequeue -H
    "Content-type: application/json" -d '{"metadata": "Sample Queue"}'

You should get an HTTP 201 along with some headers that will look similar to this:

HTTP/1.0 201 Created
Date: Fri, 25 Oct 2013 15:34:37 GMT
Server: WSGIServer/0.1 Python/2.7.3
Content-Length: 0
Location: /v1/queues/samplequeue
Description
OpenStack Messaging (Zaqar)
Readme 46 MiB
Languages
Python 96.2%
PHP 1.5%
Shell 1.2%
HTML 0.8%
Lua 0.2%