Our blog

I recently attended a seminar on cloud music services at the 2011 NARM convention. It was a spirited roundtable discussion including representatives from many major digital music services and service providers, ably facilitated by the ever articulate Ted Cohen. What struck me most about this discussion was the conflation of the subject matter, cloud music service software models with cloud infrastructure and platform services in general. The recent high-profile cloud music service announcements by Google and Amazon, companies who also offer cloud infrastructure/platform services no doubt aided this category mistake. Given the apparent confusion of many of my colleagues, I thought I’d provide some clarification of cloud terminology (and at the same time reveal my own view of the evolving cloud marketplace).

Although Google and Amazon are recent entrants (and rumor has it Apple is not far behind), cloud music services and the cloud software service model is not something new, and Google and Amazon’s music offerings in this area are not somehow predicated on inroads these companies have made in cloud computing or on the advent of cloud infrastructure or platform services in general. Consider that the International Data Corporation (IDC) defines cloud services as:

“Consumer and business products, services and solutions delivered and consumed in real-time over the Internet”

It’s not hard to think of examples of music services that go back nearly a decade that fit this description. In the last ten years I’ve come into contact with countless companies aspiring to develop ‘play anything, anywhere’ music subscription and pay for download models at least as ambitious as Google and Amazon’s recent offerings. Michael Robertson’s my.mp3.com, Rod Lord’s Mediacode, Rhapsody’s early subscription music service and moontaxi.com (my alma mater) were all developed using traditional platform and server infrastructure models.

Currently, cloud services in the Internet domain fall into three loose categories:

Software as a Service (SaaS): These are software application services that perform a specific business function consumed on a utility model. Client access is typically through a thin client or web browser based interface. Services can be accessed through a variety of web enabled devices including workstations and mobile devices. The complexities of the underlying platform and infrastructure technologies are invisible to the client and are the responsibility of the SaaS provider to manage. Platform and infrastructure systems may themselves be supported by cloud technologies and services, but are commonly based on traditional operating system and server infrastructure topologies. Google and Amazon’s recent cloud music services fall into this SaaS category as do an array of their digital music service predecessors not to mention countless other long established services in other verticals (for example Hotmail, Basecamp, Salesforce or Facebook).

Infrastructure as a Service (IaaS): Specifically server processing, file storage, IP networks, and other fundamental computing resources provided on a utility model. Managed service providers have been leveraging technologies like server virtualization, storage area networks and intelligent network traffic shaping devices to provide cost optimization and configuration flexibility for some time. In recent years some providers have developed an abstraction layer using web service APIs or browser based management consoles that provide on demand and scalable access to these resources. Notable examples of such implementations are Amazon’s Elastic Compute Cloud (EC2) and Simple Storage Service (S3) which presumably grew out of Amazon’s desire to monetize and evolve the huge latent capacity of the considerable server infrastructure they maintain to accommodate usage spikes for their own massively trafficked end user facing service www.amazon.com.

Platform as a Service (PaaS): This is a relatively new development. PaaS providers enable the deployment of client developed application code and associated configuration to an application hosting environment where the details of the platform environment are abstracted or invisible to the client. The goal is to provide on demand scalability of the runtime environment in which the client’s application and associated data schema is deployed. Google, Amazon and Microsoft all have competing products in rapid evolution in this vertical.

These cloud service categories and how they interrelate is well illustrated by the following diagram (thank you to Kate Craig-Wood for providing):

Cloud music services like those recently launched by Google and Amazon appropriately fall squarely into the SaaS category as they provide specific software functionality that is consumed on-demand through a variety of devices, and because their client-server architecture provides continuity across multiple access points. There is no significant aspect of either of these services that exhibit characteristics of (or necessitate employment of) IaaS or PaaS services. In the case of Amazon’s service they are undoubtedly leveraging cloud technologies for some aspects of the supporting infrastructure and platform layers, but this is really irrelevant to any discussion of the capabilities and market impact of their SaaS music offering.

Further, I’d argue that in the context of Amazon’s music service, since they are leveraging their own internal system capabilities and expertise, they are not really using PaaS or IaaS for their own provisioning environment. Perhaps this is merely a philosophical point, but one thing is certain, none of the recent widely publicized outages of Amazon’s cloud infrastructure service appeared to impact Amazon’s music service or any of their other consumer facing services (others like indabamusic.com and reddit.com were not so lucky). This suggests that for Amazon's own end user services they are leveraging at least a mix of cloud and traditional solutions. Who can blame them? Outages aside, there are a number of other issues associated with systems entirely deployed on PaaS and IaaS service models related to things like security, auditing, SLA conformance, and IP protection. Amazon is making excellent money selling off their latent capacity through a cloud provisioning model, and some of their business units, where appropriate, utilize their own cloud services but others don’t and shouldn’t.

What I’m trying to get across is that software service models with the characteristics of SaaS have existed long prior to the advent cloud system provisioning models (PaaS and IaaS), and that these SaaS service models don’t really need PaaS and IaaS. I'm not suggesting that the development of cloud provisioning models isn't an important and a relevant consideration for CIOs at companies with SaaS offerings (or indeed for any CIO). In today’s rapidly changing business environment, where agility is key to success and even survival, consumption of outsourced and utility services that provide true elasticity in scalability and overhead is highly desirable. Businesses whose IT strategies ignore or discount the cloud do so at their peril, but keep in mind that PaaS and IaaS services are truly in their infancy, and there are many established provisioning models (for example managed server infrastructure outsourcing, SPLA software licensing, private cloud virtualization, etc) which (properly implemented) provide the same value without the same risks.

If you’re a CIO or IT manager, I’d suggest taking a verse from Amazon song sheet: the best IT strategy for your business is one formulated through a firm grasp of the fundamental technologies and how they can best be combined and leveraged for maximum benefit, and not through a poorly conceived headlong rush into the clouds.

music.google.com
www.amazon.com/cloudplayer
Ponemon Cloud Providers Security: A Challenge and an Opportunity (securecloudreview.com)
Amazon Cloud Goes Down, Takes Every Hot Startup With It (www.businessinsider.com)
Amazon Enters PaaS with Beanstalk (www.infoq.com)
Defining “Cloud Services” – an IDC update (idc.com)

Make the easiest order sovaldi will not cause you made no problems, and it is admirable how not to sight. Here it is great for those who disburse the money.

Unassigned IPv4 addresses have nearly run out.  In February of this year the IANA assigned the last two available blocks of IPv4 addresses to APNIC (Asia-Pacific Network Information Centre).  The recent proliferation of IP enabled mobile devices coupled with the indiscriminate 'sailor on shore leave' assignment of large IP blocks to corporations in the early days of IPv4 has lead to the depletion of available IPv4 addresses.  NAT and CIDR definitely provided a significant reprieve, but ultimately only delayed the inevitable.

Simple supply and demand economics would suggest that IPv4 addresses will become a commodity in the near future as supply diminishes while demand is still strong.  ISPs and network operators will soon be charging their clients for assignment of IPs from their limited pool of available IP addresses (charging for a 'static' IP address is standard practice for end user ISP accounts, but has not been typical for business clients buying colocation space or managed services - this will change).  Moreover a 'grey market' in IPs will likely emerge in the coming months, the Microsoft purchase of Nortel IPs for a reported $7.5 million is one striking example of this trend.

The solution is IPv6.  IPv6 uses 128 bit addresses and can support up to 340 undecillion addresses (3.4×10³⁸) compared to IPv4's paltry 4.3 billion addresses.  IPv6 also brings a number of improvements from a network plumbing perspective chiefly related to multicasting and security.  Despite being an established protocol for over a decade, IPv6 adoption has been very slow.  For example, if you are using the recently launched Bell Fibe service at home (and a modern operating system) you're likely already IPv6 ready, but if you’re like the majority of folks and using traditional ADSL or cable Internet and a DOCSIS modem, you're probably stuck with IPv4 for now.  At work (unless your IT team is particularly progressive or have too much time on their hands) you are likely on IPv4 at least for Internet bound traffic.  You can check to see if you are IPv6 ready using this simple test:  http://test-ipv6.com/

So IPv4 is running out and IPv6 adoption is slow.  What should you do? 

First I would suggest evaluating your company's roadmap to IPv6 adoption.  When and how will you transition?  What will the impacts be?  Can your network support both protocols for a period of time?  If no one in your organization has the answers then you probably need put some attention on this issue and possibly get some outside help. A great opportunity to bring some attention to this issue in your organization is participation in IPv6 day on June 8th 2011. You can find out more about how you can participate in this awareness event at the Internet Society website (http://isoc.org/wp/worldipv6day/).  Second, if you have plans for significant growth in your web infrastructure in the coming months I would recommend securing some IPv4 addresses of your own by applying for an ASN and Class C IP block assignment from ARIN. 

Make the easiest order sovaldi will not cause you made no problems, and it is marvelous how not to sight. Here it is important for those who put out the currency.

151 Front Street has a long history in the telecommunications industry.  Originally constructed in 1954 it was home to the first Canadian telegraph company.  This company, verbosely but accurately named The Toronto, Hamilton, and Niagara Electro-Magnetic Telegraph Company, carried the first telegraph message in Canada sent by the mayor of Toronto to his counterpart in Hamilton on December 19, 1846.
 
The building has evolved together with the industry, was significantly refurbished in 2000, and is now one of North America's most connected buildings, boasting 9 unique fibre-optic networks, 25 diverse entrances and more than 7,000 strands of fibre.

151 Front is one of eight carrier-neutral facilities in North America and offers excellent opportunities for network peering and access to the best tier 1 providers in the world.  650x.com leverages this diversity to drive costs savings, and achieve true network redundancy for our clients.  Unlike our managed serivce competitors who mandate the use of their proprietary networks (and vastly inflate the associated costs), 650x.com is entirely carrier neutral and can provide access to any of the over 150 carrier networks available at 151 Front.

Make the easiest order sovaldi will not cause you made no problems, and it is splendid how not to see. Here it is important for those who outlay the dollars.

650x.com has revamped our website to provide updated information on our services. Like many of our clients we chose to leverage the Joomla 1.5 content management system to backend our site.  Joomla offers integrated blogging, social media integration and a variety of plugins and templates to make site management a breeze ( This e-mail address is being protected from spambots. You need JavaScript enabled to view it if you’re interested in Joomla, SharePoint, Drupal or Wordpress CMS solutions).

650x.com offers managed infrastructure solutions and data center housing for e-commerce, social media platforms, mobile marketing, SharePoint™ implementations, disaster recovery sites and more.

Make the easiest order sovaldi will not cause you made no problems, and it is marvelous how not to see. Here it is significant for those who expend the currency.