Frank Carver’s musings about software and life
“Cloud computing” is still very high on my list of things I need to get up to speed on. One trouble is that innovation in this area seems to be coming faster than I have time to experiment. Still, this article seems a useful summary of the current state of play.
InfoQ: Comparing Amazon’s EC2, Google’s App Engine and Microsoft’s Azure
Cloud computing is taking up more and more space on my virtual radar at the moment. Implementing an actual project on a commercial distributed virtualised platform is becoming a high priority for my scant spare time. It’s nice to see that the cloud market is beginning to mature, with increasingly diverse offerings and even price reductions.
Alan Williamson provides a handy calculator for calculating your cloud storage costs.
The industry in which I work involves a fair amount of content delivery over HTTP. or at least it does when the rest of the system is in place to allow people to buy stuff. As such, we have a fair amount of operation and deployment issues around keeping this working. For each new customer in a new part of the world there is a decision about how to serve content and, most importantly, where to serve it from. Making the wrong choice can lead to latency, errors, increased cost, and inflexibility.
So the announcement that Amazon are setting up a service to cater for just this situation is an interesting one. The premise is that Amazon will use their increasing global network of “cloud” servers to shuffle and cache content around the world, and use that adaptability to provide downloadable content from the most appropriate server for the end user.
Every time I read one of these “cloud” announcements, I get a small feeling that I should make some time to prototype something like our current product using these technologies. The potential pay-off of outsourcing the infrastructure could be enormous.
Technology News: Web Apps: Amazon’s New Service to Rain Content From the Cloud
As you can see if you are reading this on the blog,punchbarrel.com web site (you will need to imagine if you are using an RSS reader or some fancy blog-to-email thingy), this blog has a “tag cloud” in a sidebar. I really like the way that a tag cloud allows quick and easy access to existing posts without searching or browsing.
However, the tag cloud widget built in to the version of Wordpress I am using defaults to display of a small number of tags (around 40, as I recall.) This is a piffling amount and results in most of my low-volume tags not being displayed - largely defeating the point of tags for navigation. I did find a place in the widget PHP where this value can be overridden, and that’s how I have been showing a reasonable number of tags over the last few months.
The modification I found has one big problem. It is reverted every time I upgrade Wordpress, and so far I have ended up searching through the Wordpress code to find where to make the change again. Each release I have hoped that this obvious value will be available as a widget parameter, but to no avail. With the latest Wordpress update I could no longer be bothered to look.
So today I decided to abandon the existing tag cloud widget and look for one more suitable to my needs. The one I eventually settled on was Configurable Tag Cloud (CTC). This plugin acts a widget in the same way as the built-in one, but is much more configurable. best of all, it defaults to showing all the tags - just what I want!
As an additional bonus, and only a few months after it was pointed out to me by some of my users, I have finally fixed the CSS styling when hovering over a sidebar link, so that it no longer obscures the link text with a big opaque blue blob. Now the current link is indicated with a gentle yellow background.
Enjoy!
Obviously Yahoo is still trying to fight back against Google’s dominance of all things big and cloudy. It will be interesting to see if their approach of trying to claim some sort of ethical high ground by pitching their new cloud system at research and education pays off.
Obviously, “cloud computing” is the hot topic right now. Here’s another system which seems to be addressing the same space as Amazon’s EC2.
This project is a great idea: an open-source alternative to Amazon’s proprietary EC2 “elastic cloud” product. I can’t help it though. Every time I see the name I think of a discussion along the lines of:
Person X: I’m worried. This Amazon EC2 sounds like a great solution, but there’s only a single supplier. If they go bust, we’d be up a gum tree.
Person Y: No worries, mate. We’ll just build another gum tree, a Eucalyptus.
Person X: Bonzer!
Read more sensible stuff at Eucalyptus
The post below is another example of the way that active (powered) storage is seen as the answer to almost every computing problem. I am concerned, though about the long-term impact of such approaches.
As time progresses, and more digital data accumulates, presumably an ever increasing amount of powered storage will be required. There are essentially tow solutions to this: get more of the same, or upgrade to devices with larger capacities. Both approaches have environmental and sustainability issues. “More of the same” will require an ever increasing amount of power to keep all those discs spinning. “Swap for bigger” will result in increasing quantity of discarded hardware.
So far I don’t see anything from these cloud-storage startups to address such issues. And that’s not even considering what happens to the data if/when these companies go out of business or change priorities.
Even though this article uses the term “web 3.0″, it’s still worth reading. The idea of moving from specific “cloud computing” implementations to a more generic and pluggable approach is a neat idea. Using ruby as a scripting language for “agents” and Erlang for the infrastructure seems a good balance, too.
InfoQ: Vertebra: EngineYard’s Next Generation Cloud Computing Platform
It looks like several people are actively trying to pin down the much-bandied term “cloud computing”, and in particular, to define what it is not.