[OH-Dev] Further small (but important) improvements to the deployment

[Asheesh Laroia]

One of the long-standing problems with the main OpenHatch server is that 
it runs out of disk space.

I just did two things that should ameliorate that.

1. I moved "inside.openhatch.org" to the secondary server, 
linode2.openhatch.org

That should save 1.6GB of disk space.

2. I fixed Roundup's configuration so that it stops creating gigabytes of 
logs that fill the disk

In particular, I adjusted the configuration of logrotate so that it would 
rotate this file appropriately. Roundup will probably still make 200MB of 
Apache Common Formt logs per day, but then they'll get rotated away and 
compressed, rather than permitted to grow forever until they fill up the 
disk.

When the site has been slow or unresponsive, low disk space (and therefore 
a backlog of writes to MySQL) has usually been the cause.

Nagios is configured to monitor our disk space on linode.openhatch.org; it 
just issued a RECOVERY that disk space is (finally) fine again, with 3.4GB 
out of 16GB free.

Just an FYI that I thought some people here might appreciate,

-- Asheesh.