VMs are basically a step towards the cloud mentality. Once you're in that mentatlity, it suddenly becomes trivial to remove and add new nodes when one dies. Suddenly failure just isn't a big deal any more*.
*Assuming both scenarios have good fault tolerance where viable, eg bonded network interfaces, redundant power, RAID 5 etc
station wagon has low bandwidth, the tapes have to be written and read.
No. Bandwidth is how much you can send in one go (tapes/hdds in a car are extremely high). Latency is fairly much how long it takes you to do it. Throughput brings these together.
Hailo, a European mobile taxi network, is announcing a $17 million round of Series A financing led by Accel Partners.
Hailo, which has been operating in London since last year, is a mobile network that matches passengers and licensed taxi drivers.
This looks like a real business with real revenues that has worked in London. I wonder what this will be like when it gets to the US.
I'd go a step further to say that there were many different models of the RAZR. I have one of the higher end ones (second hand) and it easily keeps up with moden phones.
And functionality wise it does pretty well too. It's easy to teather, browse (I use opera) and run Java applications.
The OP may not like their phone, but if you read the wikipedia page, you can see the original RAZR easily makes it into the definition of a smart phone.
That would be an interesting competition.
reporter: With me now is Tom who has just booted a computer inside a computer, insiiiiide a computer sixTEEN times!
tom: Yes, it took about a year to boot the last layer of nesting and takes nearly 2 hours to run ell ess, but it was totally worth it. I'm really excited.
reporter: And what do you want to do with it now?
tom: In a couple of years, I hope to be up to 17 layers, and it's my life ambition to get up to 32.
Seriously though, if I was to do such a thing, I'd put it in a VM to begin with. So at least then I could pause and resume the VM on different hardware so that I could continually upgrade the hardware as newer stuff became available. Now what would be really cool, is to implement hibernate (or suspend to ram and dump the ram) so that the browser could be changed out without rebooting the JSVM. Now that persistent storage is sorted, perhaps that isn't too hard?
The next thing I'd do is try to automate as much as possible. Because once it starts slowing down, it would be pretty frustrating to interact with.
As soon as someone figures out a more useful way to access the internet than simply tunneling through http/s, this will effectively provide rooted-box-in-a-box. Ie stuff-all effort to get a huge number of hosts at your disposal. Persistent storage provides an easy way to resume where you left off.
US is not another availability zone -- it's a different region. There are multiple AZs per region and -- if Amazon is doing their job -- a lightning strike should not take down more than one AZ in a region.
Correct, except that it was only one availability zone that went down. That AZ maps to one of 3 different labels for different customers. For us it was eu-west-1a, for one of our customers, it was eu-west-1b. The AZs did their job exactly as they were supposed to.
Any program which runs right is obsolete.
