Cloud Computing stressing aging networking infrastructures
One of the things that drives datacenter architects and engineers absolutely mad is that networking is still a second-class citizen overall. While server processor speeds, number of cores, memory density and speed, hard drive size and other pieces of the infrastructure have increased dramatically over the years, networking has moved at a comparatively glacial pace. Before the modern push for Cloud Computing, this was a minor annoyance, only seen at certain junctions where large amounts of data needed to be sync'd across long distances. The Cloud, however, has brought this issue to the forefront of engineers' minds. If the Cloud is the heart of an infrastructure, the network is the blood vessels. Instead of substantive arteries and veins, however, we are working with capillaries.
If we compare the progress of disk sizes over the last decade from megabytes to terabytes, it would follow that in order to move that data quickly between machines we need terabit connections. Unfortunately, progress essentially stopped with the gigabit ethernet connection until recently when we saw the 10 gigabit ethernet connection hit the market. Overall, however, the amount of 10 gigabit networks in place today versus gigabit networks is tiny. Systems architects and engineers have known about this issue for a long time, but the Cloud has pushed the issue to the forefront as it stresses the current capacities of networks around the world.
Cisco has just released its Global Cloud Index for 2010 to 2015 and it estimates that global cloud computing traffic will grow by a factor of 12, from 130 exabytes to 1.6 zettabytes annually by 2015. Keep in mind that 1 zettabyte is 1 trillion (that's right - with a T) gigabytes. That is a massive amount of data that will be moving across connections that are predominantly 1 gigabit or slower. According to the report, the Cloud currently occupies 11 percent of datacenter traffic but will triple to occupy 33 percent of that traffic by 2015.
Also, by 2015, the composition and direction of datacenter traffic will shift to where 76 percent of that traffic will remain within the datacenter, 17 percent of the traffic will leave the datacenter toward and end-user and 7 percent of the traffic will go between datacenters for things like storage replication, multi-site application communication, off-site backups and Cloud bursting (federation). By 2014, Cisco estimates that over half (51%) of the datacenter workload will be within the Cloud infrastructure versus the traditional infrastructure. This is huge for private and public Cloud vendors as they must prepare to handle this massive shift of workloads to their Cloud infrastructures.
"Cloud and data center traffic is exploding, driven by user demand to access volumes of content on the devices of their choice. The result: greater data center virtualization and relevance of the network for cloud applications and the need to make sense of a dynamically evolving situation,” said Suraj Shetty, vice president of product and solutions marketing for Cisco. “The Cisco Global Cloud Index provides insight into this traffic growth and trends so that organizations can make strategic long-term decisions. We will continue to develop and release the Cisco Global Cloud Index on a regular and ongoing annual basis, contributing to ‘cloud readiness' efforts worldwide."
As a Cloud Architect, all I have to say is that Cisco, or another networking vendor better bring the terabit connectivity, and fast! When you add the total amount of Cloud traffic that will need to flow between datacenters to the growing amount of data being replicated between SANs, you quickly realize that even 10 gigabit networks are inadequate for where we are headed. Its time networking caught up to disk, processor and memory density, and that means terabit networks.
Ubuntu 11.10 – Oneiric Ocelot – Released
Ubuntu has just announced the release of the newest version of their Linux Distribution: Ubuntu 11.10 Oneiric Ocelot. With this release comes a myriad of updates and bug fixes along with further development of the Unity UI. The server version aims to take hold of the Cloud with a plethora of built-in cloud tools. A brief excerpt from the official release letter is below. You can download the newest version at http://www.ubuntu.com/download/ubuntu/download
"There is nothing like a dream to create the future." - Victor Hugo
The Ubuntu team is pleased to announce Ubuntu 11.10, code-named
"Oneiric Ocelot". 11.10 continues Ubuntu's proud tradition of
integrating the latest and greatest open source technologies
into a high-quality, easy-to-use Linux distribution.For PC users, Ubuntu 11.10 supports laptops, desktops and netbooks
with a unified look and feel based on an updated version of the
desktop shell called "Unity", which introduces specialized "Lenses".
Finding and installing software using the Ubuntu Software Centre is
now easier thanks to improvements in speed, search functionality
enhancements, and usability improvements. Aside from updates
on the performance side, it's also more aesthetically appealing.Ubuntu Server 11.10 has made it much easier to provision, deploy,
host, manage, and orchestrate enterprise data centre infrastructure
services with the introduction of "Orchestra". The Juju technical
preview allows service developers to describe the deployment and
scaling requirements of their applications, in order to simplify
and enhance the dialogue between developers and operations teams.
For those working on the ARM architecture, a technical preview is
also provided for the ARM server.Read more about the new features of Ubuntu 11.10 in the following
press releases:http://www.canonical.com/
content/transforming-home-pc- ubuntu-1110
http://www.canonical.com/content/client-cloud-ubuntu- 1110-sets-pace-business-it
HPC Cloud Service for Researchers coming soon
One of the areas which pops up from time to time in Cloud conversations is that of High Performance Computing (HPC). The HPC space, perhaps undeservedly so, is a niche market usually comprised of researchers from sectors such as biomedical and aerospace. This are has always interested me because I have a personal affinity toward science and discovery and I feel that often research projects don't get the full amount of funding needed to be successful. Personal interest aside, HPC in the Cloud is just one more area where large multi-tenant farms of commodity hardware could be put to more efficient use for the greater good.
When looking at use cases for HPC in the Cloud, one must look at the individual workloads that will be processed and make an assessment as to what level of true scale is needed to accommodate multiple workloads. The level of parallelization of the software which will run in the Cloud is of high importance as is the speed at which the Cloud can execute operations. While some software written for traditional on-premise HPC clusters can scale quite easily in a horizontal fashion, some software was simply not written to scale beyond what was probable in terms of on-site equipment. Thus, the first step is to ensure that future software, if not already written as such, should be written to scale horizontally indefinitely. Software that can only scale to a certain 'width' can still be run in the Cloud, but without several other pieces of software with similar limitations sharing the same 'tenant space', the workloads will be imbalanced. This is simply an issue of categorizing the horizontal scalability of an app and ensuring that resources are aligned as much as possible underneath similar types of applications.
Once we have sorted out how to balance the workloads across the multi-tenant platform, then we can start to process the workload in batches that make 'resource sense.' It goes without saying that having workloads 'trespass' onto areas where resources are not geared for best performance of that workload will not be the most efficient, but it does offer some flexibility if one type of application workload or the other has terminated early and the underlying resources are free to be re-assigned.
Now, with all this said, it sure does help when a vendor steps up to make the transition easy for researchers and scientists from on-premise equipment to HPC in the Cloud. One such vendor, Logicalis, has done just that for researchers in the UK. The truly remarkable part of what Logicalis has done is allowing many entities to pool funding and purchase high performance computing power. Think of it as a Public Cloud, similar to Amazon, but whose principal clients are researchers in need of HPC infrastructure. It really is a brilliant move and I hope that it is mutually beneficial for Logicalis and the researchers involved. Having a relationship like this work in the field of research and HPC just lends more credibility to Cloud Computing overall.
You can read more about the Logicalis announcement here and you can read more about Logicalis here.
