By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Amazon debuts new instances
Apparently encouraged by the gobbling up of high-memory capacity virtual server instances, Amazon Web Services (AWS) has announced a new pony in the stable: the High-Memory Extra Large Instance (m2.xlarge). The cloud provider's latest creation comes with 17.1 GB of RAM, two virtual cores (equal to six and a half standard EC2 cores) and 420 GB of instance storage. At $0.50 per hour, this is definitely a budget instance designed for high-memory cache applications.
This will fill demand quite neatly between users of the standard Extra Large Instance at $0.68 per hour, which has 15 GB of RAM and five cores but gobs and gobs of storage and I/O, and the Double and Quadruple Extra Larges, which double both the price and the available RAM. Even in these futuristic times, 17 GB of RAM is a heck of a lot of memory to play with, and AWS has no doubt accurately forecast the product niche and delivery here.
For those of you playing the "Guess what Amazon's got under the hood" game at home, 18 GB of RAM times four equals 72 GB and two cores times four equals eight CPU cores. You could buy that off the shelf for about $6,000…not that Amazon is paying full retail. (Perish the thought!) Four users on one box make EC2 $2.00 per hour, or $17,500 per year, a gross profit that clocks in at around 290%.
So in case anyone was wondering if Amazon was making any money on cloud…
AWS has also announced reserved instances or pre-bought capacity for Windows server instances, potentially bringing a year's worth of Windows CPUs under the all-important $0.05 per hour sweet spot. Actually, no one knows what the sweet spot for Windows servers is, but they just keep following Linux instances around like a lame friend from college you totally don't want to hang out with.
gear6 adds more features to its AWS memcached service
In relatively related news, for-profit open source firm gear6 has announced that the AWS version of its memcached service can now scale out live RAM caches on demand.
Dynamic Scaling is said to "eliminate a major issue with memcached -- its inability to maintain cache data while changing cache capacity, resulting in unnecessary site degradation or downtime." We don't know if anyone thought that was a major issue, but it's not a bad add-on for the busy cache proprietor. And right on time for the m2.xlarge!