2,752 karma · joined August 13, 2009
- PostgreSQL for the user database since it maps nicely to django and sqlalchemy models
- Redis for caches and non persistent data
- Riak for everything else
We love the fact that it is easy to scale, that bitcask is safe (it writes to disk:-p) and easy to backup (rsync or cp is enough).
We already have one alpha user that seems pretty happy with it.
It will work without an upgrade.
The situation got much better since 1.8 and especially 2.0.
The main reason zeromq is so awesome is that we can use it as an rpc protocol for almost everything.
If we start seeing a recurrent pattern we'll try to automate.
- 0.7.65 is the version currently used in ubuntu lucid. We use it only as a proxy.
- On services like python-wsgi we currently use 0.8.52 and we will soon upgrade to 0.8.54
- On the ruby-passenger service we follow the nginx pulled by passenger
So it's always dependent of the context. The idea is to keep the stacks stable and secure at all times.
Once we have tested and approved an upgrade we create a new revision of the corresponding service and deploy it across the platform. If the upgrade requires downtime we schedule a rolling maintenance window and notify users.
On caching:
We don't cache by default (but this could change). Users have the option to add caching services like varnish.
For example:
- for django you will have an easy way to add more instances
- for mysql more slaves
- for celery or resque more workers
http://www.tug.org/mactex/morepackages.html http://mirror.ctan.org/systems/mac/mactex/mactex-basic.pkg.z...
It would be a lot of work to abstract the js core.
Edit: I notice that a lot of people think ZFS is a cluster/distributed filesystem. I don't understand where they get this idea.
It's all projects where I need to connect different protocols together. Like amqp & websockets for example.