Leave a Message
Disclaimer: The information in this site may include opinions or views of
the blogger which, unless expressly stated otherwise, are not necessarily
those of Gen-i or any associated entity. Please contact Gen-i directly in
respect to any questions or queries arising from any comments made on this blog site.
Visit www.gen-i.co.nz or
www.gen-i.com.au for contact details.
angelina valentine and rachel starr angelina valentine facial angelina valentine movies club angelina valentine angelina valentine long spear angelina valentine footjob angelina valentine oral angelina valentine foot angelina valentine kimber james angelina valentine xxx free angelina valentine in latex
Well, the Christchurch Cloud Camp on Friday the 23rd was a good event with a lot of great discussions, and a real mix of local corporate and start-up companies present. The afternoon was spent discussing some specific topics, which were split into a more technical group, and a business focused group. Some great advice from Xero for start-ups as well. Despite my roaring Flu bout, and with some assistance from ‘pharmaceuticals’, I led off with the first lightening-talk which focused on “The Cost of the Road to the Cloud” and a quick last minute discussion on the Cloud security and ‘Identity’ challenge that follows this subject usually. I chose these topics, as in my many discussions with customers and other consultants in the industry, many many people forget that the most variable cost with the ‘International Cloud’ is the network data costs. I thought I would summarize the conversation thread that I originally mind-mapped in this Blog, with some post Cloud Camp footnotes…
The International On-the-Road cost:
- To get to the cloud, you need a network of sufficient bandwidth – this amount is an educated guess at the start
- Ok, so what does it cost roughly… well, it averages at about 14c/GB for traffic to the cloud, and 21c/GB for traffic out of the cloud (MS in Asia is higher – over 50c/GB out)
- Ok, so what if my business has seasonal peaks or campaigns, and my traffic increases by a factor of 10? Or 100? … or 1000 times the quite periods? … and how do I apportion this cost back to the business unit responsible (if I can)?
- You need Quality of Service (QoS) to differentiate your international application traffic classes as well? … sorry… come back in 2 years…
[Footnote: These comments refer squarely to International bandwidth costs - greater control and predictability of cost and BW options are available from an NZ only Cloud connection]
Some Scenarios… backing up my data:
- OK, so I have my 2-tier app & DB hosts in the cloud, and I have some paying users. Now my company has been told by one of the ‘big-5’ auditing consultants that we need local (e.g. NZ, onsite) nightly backups at all times for compliance and to keep the financial risk managers happy (e.g. the bank, in case my cloud provider becomes vapour).
- Every night we now backup across the network, using data extracts, because differential backup snapshots aren’t an option (e.g. traditional SAN and tape backup solutions). This means big chunked extracts from the app database. During BAU, this is about 5-20GB a night, unless I develop my own ‘sophisticated’ differential DB extraction solution
- At average international network prices, this is about $1 – $10 per night, depending on your international cloud provider
- Ok, so now with BAU, that’s about $364 – $3,640 per year, just for backups resident in NZ, and I still need NZ network, storage and archiving infrastructure for all this once it’s back here
[Footnote: Traditional SAN snapshot technologies used with large database backups aren't available for hybrid cloud scenarios, so you'll need to consider how to back your datasets up locally with your cloud provider. Even alternate Backup-as-a-Service (BaaS) providers don't solve this BW cost issue. Of course, this is less of a challenge using an NZ based cloud provider with a quality network ]
Some Scenarios… campaigns and peaks:
- So for BAU you average 1000 user sessions a day, and they typically navigate around 8 pages of content – circa 2MB of outbound network traffic per session… that’s about 50c of network traffic for the day
- Now we run a campaign, and the email, twitter, and facebook link-backs drive the traffic to 90,000 NZ sessions a day – that’s about $450 a day, and hit rate of about 2% of our population BTW
- But now we’re a global company, desperately trying to gain global market share for our niche product. Our next campaign generates over a million hits – that’s about $5000 a day of traffic! $35k over the ensuing week – Ouch! Didn’t budget for that in the campaign. Time to look at a better Content Delivery Network (CDN)…
- So the network cost turned out to be far more variable than the CPU or storage costs from my cloud provider, and we couldn’t budget for this at the start of our project (just inexperience)
[Footnote: This is a slightly contrived scenario, but you get the message - the peak international network costs can easily exceed the best local network solutions! ]
Future proofing – IPv4 addresses are running out, you say?:
- Ok, so we look to our cloud provider for their solution to the diminishing IPv4 addresses, and their IPv6 readiness and roadmap for global connectivity to your host… Hmm… can’t seem to find anything on their web site – time to send some enquiry emails…
- What’s this? IPv6 is still in ‘experimental’ stage? A future feature? Ahh… when in the future will that be, you ask?
- Uh oh! So I can only do this on my self-managed VM (if my cloud provider can offer this), and I have to configure my own Teredo, ISATAP and 6-to4 gateway(s) somewhere? Now I have to employ a network consultant?
- What about my East Asian manufacturer who only has an IPv6 enabled ecommerce web host? (a reality today BTW) Why am I now building a network on top of my cloud providers network?
[Footnote: This is the great 'elephant' in the room at the moment for all Internet connected hosts. Your network provider must have an IPv6 strategy and roadmap that they can articulate to you ]
Other subtlety’s…:
- Don’t forget, yes, you have to pay for NZ local network access to the Internet as well, and that the cheaper it is the more contended the bandwidth will be with other customers, and yes, there will likely be a pre-agreed monthly cap on this that looks ugly when you exceed it!
- Did you know Google are tweaking their ranking algorithm to measure page load speeds right now?
- Want to (Google) rank above your global competitors? What are your options?…
More bandwidth
Use HTML 5 and Web Sockets
Use cloud providers with smart caching and Content Delivery Networking (CDN) solutions
[Footnote: There are many subtlety's with network attributes and behaviors, keep it simple in the initial focus and think about the 'path' between your cloud app users and where the application is executing ]
Now what? … Cloud Security…:
- So you thought you cracked the network cost model, after the first campaign ‘peak’ made your CFO choke on his tea and scones…
- Now the company needs more real-time integration with NZ back-office systems and partners in your supply-chain and your ‘Big-5’ auditor is asking about PCI DSS security and SoX compliance because you needed a US office to do business there to be credible
- Wait – your cloud provider uses username, password, and a ‘token’ for identity? How do I synchronize and manage these from my existing AD management tools? I can’t? @#$%!!!
- My data is sensitive, and we used 2-factor authentication for our non-cloud version of our app – who can provide this? What? No-one currently? Unbelievable !!!
- Hang on, now I’ve done some due-diligence, so I have a choice of doing some bespoke integration for Single-Sign-On (SSO), or buying a $100k+ Identity Management solution, and paying a consultant to configure it. My CFO is going to choke again…
[Footnote: The simplest of Enterprise Architecture deliverables suddenly become a challenge in the hybrid cloud model - don't underestimate the hidden complexity and cost!]
Wait – we didn’t budget for these kinds of network and security issues! :
- There is a moral here…
- Cloud solutions require different thinking about the cost (and service) models, and this requires a different skill set from traditional ICT projects and roles
- Plan, plan, and plan again
- Bring in consulting resources who have done this before – you’ll be better off for it
- Really ask yourself if you need to use an international cloud provider to accomplish what you need in the first instance
- Take the time to do due diligence on all the providers out there, and make sure their IaaS and PaaS is aligned with your chosen technology roadmaps – put money aside to do this, and consider it part of business ‘risk management’
[Footnote: While these sound like common sense, it is surprising how many people don't 'tick' them off! ]
So on all that note, putting careful due-diligence into how you design and provision your network road to your cloud provider, or at least understand it’s cost model, service levels, and behaviors, is paramount as part of your (or your customers) cloud solution delivery.
Next time I’ll focus back on some of my other favorite subjects – “The differences between PaaS providers, emerging new software languages that are Cloud-ready, and market influencers driving acquisitions in the cloud start-up ecosystem”.
I really mean that cloud is become necessary. for example, a normal shop owner wants to see his camera from his home computer, cloud is needed. Besides all other technology, I personally feel that ubuntu is one of best and secure for normal man need and requirement.
CloudCamp Wellington
I attended my first Wellington cloudcamp on Friday afternoon. As always, it was very well organised by Ben Kepes and Ruth McDavitt. They provide a great, relaxed atmosphere that’s conducive to open discussions. Unfortunately I could only stay for the first two hours as I had other commitments that afternoon, which is a shame as judging by what I did get to hear I’m sure I missed some useful stuff!
I gave the opening lightening talk and decided to share some of the sentiment we at Gen-i have been hearing from our clients around cloud computing. Suffice to say there is a lot of interest, as industry and media hype has not been short in extolling the potential benefits of ‘the cloud’.
Clients see lots of potential benefits, many are specific to them and their ICT environment, but the most commonly discussed reasons for considering cloud services include cost savings, increased business agility and flexibility, and the availability of mature enabling technologies.
Recessionary pressure has led to them being open to new ways of procuring ICT, and the opex based consumption model of cloud is ideal for this. Not to mention the significant shared-cost benefits that multitenant services provide. Closely linked is the ability to access services on an on-demand per-seat basis means they can reduce or increase the quantity or type of computing resources as required. This is a radical change for those who have looked at ICT as a sunk investment, and an investment that needs regularly updating at that! A CFO in the construction industry described it this way: ‘when I invest in a concrete mixer I see a return on that investment, but when I invest in a new server, I don’t see the same return. It’s not our core business, and to us it makes sense to procure ICT as a utility, much the same as we do electricity’.
Although many of the enabling technologies (such as broadband, virtualisation and web services to name a few) have been around for some years, their maturity and availability make the business grade delivery of cloud services a reality. Still, there are many concerns on the minds of our clients, such as on-boarding these new services and integrating them with existing systems. A ‘greenfields’ approach is ideal, but very rarely practical as there are so many technical dependencies to linking the often myriads of workloads and work processes that incorporate modern IT. And the issues are real. One company I heard about recently had assesses their application portfolio to determine how many could be migrated from on-premise servers to the cloud. A technical assessment determined 50% could be transitioned to a cloud environment, but when a business process risk and cost assessment was completed, they found that moving less than 5% was really practical. The costs and risks were deemed too high.
So, is the industry hype wrong? Is it impractical to realise the benefits that cloud computing is touting? It’s still early days, so it’s just not practical to make assumptions just yet, however one thing is for sure; service providers and system integrators will play an increasingly important role as purveyors of business outcomes through ICT.
The ICT industry is being transformed, in a similar way to how the shipping industry was transformed 60 years ago by the introduction of containerisation. The advent of containers led to a shift in focus to supply chain logistics as customers became less concerned with who was carrying their goods, and more concerned with reliable, safe and consistent delivery.
That’s exactly what is happening to ICT; clients are becoming less concerned with the type of ‘containers’ (vendor technologies) and more concerned with the end-user experience and business outcome or ‘logistics’; integrating the various vendor components and offerings to provide reliable, flexible and cost effective ICT services.
The SI’s who get to grips with this and excel as ‘logistics’ providers of ICT will be well placed to make the potential benefits of cloud computing a reality.
Cloud camp was interesting, my presentation seemed to be well received, at least I didn’t get flamed on Twitter
fierte: Jonathon talking about my sort of company at #cloudcamp – Packaging cloud apps for customers and providing unified service
michaelharries: RT @michaelharries: Johnathan of geni-i – practical cloud concerns and cloud traction. #cloudcamp http://twitpic.com/16ild3
The general consensus seems to be that there was still a great deal of miscomprehension in the market place as to what the cloud actually was, and general fear on issues trying to exploit cloud offerings.
In the breakout sessions I attended there was quite good conversations around compliance and legal issues, data portability and emerging opportunities in the ANZ market.
One of the comments that Ben made around legal issues was that in general, the conversation in Australia was more relaxed than the same discussion that was had in NZ. People here recognised it as an impediment to the future of cloud, but not an insurmountable issue.
The data portability discussion tended to go around in circles, however to me the clear message was that there is an excellent opportunity for a company to carve themselves out a niche specialising in migrating data to, from and between cloud providers, giving consumers greater mobility within the environment.
The final breakout session I attended was around opportunities in the cloud, this was largely focused on start-up companies producing new products and the challengers they face. We had a lady in the forum who had previously worked for Gartner who suggested that in general Australians were risk adverse so it is hard to get venture capital for these sort of products, whereas people in America, particularly in California, were more willing to invest in start-ups. The use of social media was also raised as a great way to break into the market and get wider acceptance of products on a global scale.
On the whole it was great to talk to others in the industry and to hear their perspectives on where the cloud was heading, although there was nothing revolutionary raised in the discussions. I think the most valuable part of the event was the breakout sessions, talking with other people on their experiences, problems and successes.
Being my first ever CloudCamp, not sure how to attend an un-conference, I was apprehensive about flying jargons and mocking experts. But I was pleasantly surprised to see a lively bunch of people who had come together to share ideas, build on others and in general grow the body of knowledge. Ben was a very good host and moderator, who knew exactly when to talk and when to shut up – which can be a problem for many a host, at some other events I have been attending…
It was a very fruitful session for me with quite a few passionate debates over data ownership, privacy issues, Google amassing private data and selling for profit and so on – notable parties being Ryan from Intergen, Paul from Gen-i and Ram, from FieldLinx.
In short, my take from the event:
1. Cloud computing needs to be better defined as a concept. Doing this will help in clarifying many issues which were discussed such as who owns the data, who can ask for the data, how can government contribute through a data policy and so on…
2. Law or Policy is always catching up with new technology and that is a given. There is no point in discussing the “why”, but in short, government has too many age “old” issues like unemployment, infrastructure development, education, healthcare etc to take care of than spend time and money in investigating the feasibility of new technology and making law to thwart the future crimes of some data pirate!
3. Being the first CloudCamp for me, from a MagnumMac (my employer) point of view, I feel that this will be a golden opportunity to start a conversation with our clients and having a clear understanding of what is it, that they would need to facilitate implementation of any such strategy? Also, knowing about it and knowing how to make it work are two different things so I would not go as far as to say that there is a concrete action plan.
Well, in conclusion, it was great to express my cloudy thoughts and meet some like-minded people at this un-conference which was CloudCamp. Hoping to drift to next year’s event as well!
Nice write up Paul…very nice summation of the entire experience!
Thanks Swapneil. Glad you enjoyed it.
I’m really heatened that you see the opportunity that it presents. Good luck
This is my second New Zealand CloudCamp (see my informal presentation here) , I was generally pleased to see a whole lot of new attendees as well as some from the first event. The thing I noticed the most was a big increase in the numbers of developers / start ups at this event. I think this reflects the general view that Cloud represents an opportunity for growth and let’s face it, everyone wants a piece of that action.
But having said that interest in cloud computing is improving. There is clearly a big job to do increase peoples and companies’ knowledge and understanding of it. This fact is born out in discussions we here at Gen-i are having with our customers and also in the hot topics of discussion at CloudCamp.
For example we had a lengthy discussion in the panel about the legal and compliance implications of cloud. For some this was a very old discussion (the same as when the internet became popular), but for others some of the issues around SLA, data ownership and data sovereignty appeared new.
CloudCamps aren’t about addressing these types of issues per se, but I was struck but the fact that the most common course of action appeared to be to ignore this because it appeared unsolvable. To me this stance puts you and your prospective cloud provider at risk. You, the customer are at risk because compliance is your job, but equally if your service provider can’t come up with an economically sound way of addressing it, they possibly won’t be around for the long term.
Another hot topic that came up in both lighting discussions and in the break outs was the complexity associated with getting into and managing cloud computing. Ben Kepes (CloudCamp facilitator, his review of CloudCamp here) really called this out. He said the almost universal proposition from cloud vendors that you can chop and change providers is a myth.
The reality is that once you are with a provider you are almost certainly going to stay with them. This was reinforced by discussion in the architectural stream about how to migrate apps into the cloud and how to move between clouds that went on. I only listened in briefly to that track so can only summarise it but I think it’s accurate to say that there are technical solutions to both questions but none of them are easy. Prospective buyers out there should know this from the get go – it’s not as easy as people make it out to be.
The next major meme that I picked up on was how the entire cloud industry (apart from some noticeable exceptions) is struggling with a robust business model. This train of thought started with Darryl Burling of MS discussing what he thought the longevity of different business models were. It was further reinforced in a break out group in which pricing, partnerships and propositions were discussed. Interestingly Freemium models such as that used by Google got a short thrift (see this post free isn’t a business model) .
“The only time success comes before work is in the dictionary” – Vidal Sasoon
As this discussion went on it became clear to me that there is an almost unreal expectation that Cloud computing (and SaaS in particular) somehow magics up new economics. This just isn’t true; you cannot ignore sound business principles. You have to know what your customers want, what they will pay for that value and how you can economically make and deliver that service. The internet and cloud hasn’t changed any of that. If you are looking to buy a cloud service, I would strongly council you to think about the prospective providers business model, because if its looks to good to be true…well maybe it is.
In the first Cloud camp, I voiced the opinion that cloud computing was nascent here in New Zealand. I am really pleased it appears that there we have moved on quite a lot. But we have our challenges. As with all disruptive innovations, there are both opportunities and threats. I fear we aren’t making the most of the opportunity (tackling the world, instead of just New Zealand), and we moving too slowly, and this could seriously impede our ability as a nation to continue to innovate through ICT. Finally we appear to be acting in isolation, lots of small companies in motion and none with the wherewithal to make a big step.
To me we urgently need to support each other and build frameworks and technology platforms to help us in making it on the global stage… there is no reason why a kiwi company couldn’t have built Salesforce.com for instance (in fact they do a lot of development in Sydney). Cloud eliminates the scourge of distance and we should be making the most of that…
My final point and something that I definitely wasn’t expecting was the negative perceptions towards Google. Comments ranged from how Google tends to ruin revenue pools with Freemium; through to Google not really understanding business needs to a general distrust about how much information Google can collect on you. This thinking was quite a turn around from last year’s CloudCamp and reflects other views that have been showing up in the wider tech community.