Great news on the VPS front though – I’m currently using Memset, who I’m pretty happy with, and who provide a lot of server for the money, and have generally been as responsive as I’d want them to be.

Good to see that ecohost have updated their site though – lets hope there are more options available for us before spring arrives!

The really good news though is that apparently VPS is in the pipeline so maybe your search will soon be over.

Nice to hear from you again Steve!

Well… I’m tracking exercise through Runkeeper here, and I’ve cut out alcohol since coming back to London, which leaves a few less factors to track.

So far, I’ve been going out a couple times a week, but managed to keep to the 8 hr rule.

So far, I’m not noticing a huge difference – I don’t feel much more productive, and to be honest, I miss the extra waking hours more than I miss the sleep, if there’s no real difference at the end of the next week, I’ll think about seeing how I get by for couple of weeks on 7 hrs again after the month.

I’d probably throw in caffeine and alcohol consumption, social interaction and entertainment as independent measures as well.

And if we are going to get this carried away you might as mash in weather data like temperature, sunshine, and barometric pressure.

To be more precise Rake do have CLI args parsing, but it’s such a piece of shit that anyone actually use it. And it’s pretty buggy as well.

Capistrano tend to have some real troubles with Git when deploy_to & repository is on the same machine, so I also wrote http://github.com/botanicus/git-deployer which works – and here the git magic comes – on post-receive hook. So when you create a project, you run some tasks for pushing a hook to the remote server and then when you do git push myserver master, than the hook run and can migrate DB or do whatever. Pretty handy, it can deploy to more servers, but it’s definitely more lightweight tool than Capistrano. BTW – git-deployer is in quite alpha stage so far.

My hosting provider has a HOWTO page on this which I found useful: http://rimuhosting.com/howto/memory.jsp

“The reason we can run our data center on 100% solar is because of everything we have done to reduce power. Our desktop computers are thin clients that only use 5 watts of power. Our AC uses 200 watts of power. It is water cooled by water we have collected in large tanks behind the data center. We are 100% virtual which means we can put 70 servers inside 1 server using the power of VMWare and NetApp. We are running a total of 10 physical servers that handles around 700 total servers. PG&E a large power provider in Northern California went from 3500 physical servers down to 25 physical servers using the power of VMWare.

We are in the process of hooking up our small wind turbine which is
powered by the ac fan. The wind turbine uses the power of the ac fan by
sitting in front of the fan. Since it turns 24 hours per day the power
we will get will power of desktop computers inside our data center. The
desktop computers will be on a totally separate circuit than the rest of
our power.

We have solar panels on order that will be going in to separate our ac
power even though it is so low. The solar panels will give our ac a
total separation from the rest of our data center. We are building a
duct work that will sit under the solar panels. The cool air generated
from the outside exhaust of the ac will be ducted to the bottom of the
solar panels. This will jump the efficiency of the panels by 15 to 20%.”

Thanks Kate for the link to the cloud article, very useful. The reason that ‘chasing the moon’ hasn’t appeared anywhere yet is because Tom just mentioned it to me in a phone call, not on his blog. Not sure if it’s a bona fide phrase anywhere yet, but let’s make it so… I think we can have support all around the world – people will just have to work night shifts!

Re. Aiso what we really need is better metrics to define what is meant by the claims – I took 100% solar to mean that their entire data centre is ‘off-grid’, i.e. not using any brown electricity off the grid. Their explanation at http://www.aiso.net/technology-network.html could be a lot clearer.

I’d like some proper clarification from them and others making similar claims so that providers can be compared more readily. Please note that I am not suggesting we create another bloody ‘eco-label’, just that the relevant clean vs dirty energy data is easily displayed and queried. That’s partly what AMEE (my employer) is for.

I’ve invited Aiso to post here if they want to respond to the above comments.