For some time, my office PC had suffered from multimedia niggles. Some were relatively minor, such as Facebook and Twitter videos failing to play, but others were more problematic. For example, some work-related videos on YouTube would play, but others would not.
The straw that broke the camel’s back was when I discovered that I couldn’t record calls made within Skype for Business (“the app formerly known as Lync”). With a couple of typically-cryptic Windows error messages as my starting points, I started digging, and it wasn’t long before clues emerged pointing to what we used to call the Windows Media Player, or WMP.
Those who’ve installed Windows more often than I have will already have guessed it: this PC is running not Windows 10, but Windows 10N. The N versions of Windows lack media playback and were produced in response to the 2004 European Commission antitrust ruling which found Microsoft guilty of anti-competitive behaviour in bundling WMP as standard.
The problem is that in taking WMP out, it seems that Microsoft took out a lot more than just a media player – it took out a chunk of stuff that is pretty much a necessity these days. It’s a bit like saying, “You don’t want Internet Explorer? Ah, that must mean you don’t want the Internet – right, we’re removing all the networking…”
Fortunately, the fix was relatively easy: you can download the missing bits from Microsoft’s website for free. A few minutes and a reboot later, and all that media stuff was working at last.
Building a system or an app to be truly modular
Of course the world has moved on a lot since then, as has Microsoft. And perhaps the original N versions of Windows XP were just a quick fix in response to that EC ruling.
However, it seems extremely odd that Windows 10N should still be in this state over a decade later. Sure, you don’t want WMP, Skype, the browser and so on to all have their own individual multimedia subsystems, when they could all use the same one. But you don’t do that by making everything else work through WMP – you modularise it properly.
Either you make the absolute minimum bits that everything needs a core part of the operating system, and WMP is then one of several programs that can call those, or you flip it around and use APIs that allow other applications to call WMP – or any other media player – when they need video playback.
This is a fairly coarse-grained example, but as modularity grows more popular in many areas of software design, whether it’s via containers, hyperscale computing, microservices or some other paradigm, it’s going to be interesting to see whether it pops up again – or perhaps, how often it pops up again.
Edit… And now the update to Windows 10 v.1803 has broken the WMP add-on, necessitating a fresh reinstall. That’s clever, Microsoft – way to go!!
Timed to coincide with the RSA security conference, Microsoft has announced Azure Sphere, a new solution for creating highly-secured, Internet-connected microcontroller devices. But there’s a bit more to it than that. Enabled by the company’s newly found ‘growth mindset’, Microsoft engineers can happily use the best technology for the job which, on this occasion, just happens to be Linux.
More than 9 billion microcontroller-powered devices are produced each year according to Microsoft’s blog post, and these are increasingly being connected to the Internet-of-Things (IoT). Similar to, but less sophisticated than, system-on-chip (SoC) designs, microcontrollers are used in many of the products and devices that surround us. These range from automobile engine control systems and medical devices, to building controls, appliances and children’s toys. The size of this market dwarfs that of the PC and mobile markets combined, and Microsoft clearly sees a lucrative opportunity if it can establish Azure Sphere as the protective ‘mother ship’ of the world’s smart things.
Azure Sphere, a secure home for smart things
In a world where so many ‘things’ are connected, almost anything, it seems, can be disrupted. It therefore follows that every connected thing needs to be protected. This is the ethos that underpins Microsoft Azure Sphere. Azure Sphere includes three components that work together to protect devices at the edge of the network: Azure Sphere certified microcontrollers, Azure Sphere Security Service, and Azure Sphere OS. Unpacking the announcement reveals a couple of surprises. First, Microsoft said it will license its technology to chip makers on a royalty-free basis. Next, the new microcontroller operating system is based on a custom Linux kernel, not Windows.
The Linux kernel has formed the basis of embedded operating systems and microcontroller-based products for many years, so Microsoft developing its own custom Linux kernel isn’t technically remarkable. However, it does jar somewhat with the Windows 10 IoT everywhere proposition of 2015, although the company was just starting to look at microcontrollers back then. Azure Sphere is still in private preview, but the first Azure Sphere chip, the MediaTek MT3620, is expected to ship in volume this year. The first wave of Azure Sphere devices are expected by end of year and, who knows, Microsoft might even ship a new device of its own. The much talked about Andromeda perhaps?
The seven properties of highly secure devices
Microsoft researchers began exploring the microcontroller-powered devices market in 2015. Then, in March last year, they published a research paper, The Seven Properties of Highly Secure Devices, that could have, theoretically at least, prevented the Mirai botnet attacks of 2016, when an estimated 100,000 compromised IoT devices took down several high-profile websites using DDoS attacks. In short, the paper concludes that a redesign is necessary if we want our IoT devices to be safe and secure. The paper also details a proof-of-concept project with MediaTek, a Taiwanese company that develops SoC for mobile devices, home entertainment systems, network and connectivity equipment, and other IoT products.
Here’s the list of properties, and their tests, which Microsoft deems critical if we are to have highly secure, network-connected devices:
- Hardware-based root of trust: Does the device have a unique, unforgeable identity that is inseparable from the hardware?
- Small trusted computing base: Is most of the device’s software outside the device’s trusted computing base?
- Defence in depth: Is the device still protected if the security of one layer of device software is breached?
- Compartmentalization: Does a failure in one component of the device require a reboot of the entire device to return to operation?
- Certificate-based authentication: Does the device use certificates instead of passwords for authentication?
- Renewable security: Is the device’s software updated automatically?
- Failure reporting: Does the device report failures to its manufacturer?
If we think about the damage caused by malicious botnets over the past decade, each comprised of tens-of-thousands of vulnerable Windows PCs, then it’s easy to see why these principles have become a bit of a thing at Microsoft. The company clearly wants to avoid the mistakes of the past, and while nothing has been explicitly stated, it would be a comforting thought if Microsoft were to apply the seven tests listed above to all its device efforts going forward. Having said that, I’m struggling to see how today’s Windows operating system model would ever make the grade. Maybe we’ll hear how in the coming months.
From security chumps to security champions
Continuing his mission to establish a ‘Digital Geneva Convention’, Microsoft President and Chief Legal Officer, Brad Smith, used his RSA security conference keynote to remind delegates of the perils of not updating and patching the Windows operating system. He pointed to last year’s massive cyberattack when, on May 12, more than 300,000 computers running Microsoft Windows were affected by WannaCry ransomware. Mr. Smith also highlighted the NotPetya cyberattacks that targeted Ukraine businesses and institutions, affecting the Windows-based systems of banks, government departments, newspapers and energy companies. These events were not cataclysmic, this time, but they do show us how the disruption of non-patched or out-of-date computer-controlled social infrastructure can affect businesses, governments, and millions of people.
Championing cybersecurity, Brad Smith also used the RSA conference to announce the Cybersecurity Tech Accord, a public commitment among 34 IT companies to “protect and empower civilians online and to improve the security, stability and resilience of cyberspace”. In essence, those companies signing-up to the accord pledge not to assist governments in cyberattacks. This list includes key enterprise IT suppliers, such as Arm, BT, CA Technologies, Cisco, Cloudflare, Dell, HP, HPE, Microsoft, Oracle, RSA, SAP, Telefonica and VMware. However, notable names not yet on the list include Apple, AWS, Google, IBM and Lenovo. We’ll have to wait and see if the accord has any real impact on the escalating political tensions associated with state-sponsored cyberwarfare, but it’s at least a start.
Nadella sets Microsoft free with ‘growth mindset’
The Windows operating is still a very important asset to Microsoft, and will be for many years to come. However, Satya Nadella has made it very clear where the company is heading: Intelligent Cloud and Intelligent Edge. The departure of Terry Myerson, Executive VP of the Windows Devices Group, signalled the start of this new epoch, and this week’s announcement, choosing Linux over Windows, provides yet another glimpse of the cultural change and ‘growth mindset’ that Nadella is driving at Microsoft, one in which nothing should be taken for granted and where employees must always be willing to check their assumptions as new data is revealed.
Microsoft is clearly changing its approach to security as with many things these days, but it still tends to blame the ‘bad guys’ and laggard IT departments for the woes caused by rogue Windows-based computers. The company has made billions of dollars from sales of its software, so we shouldn’t feel sorry for it when it has to spend millions of dollars cleaning up the mess caused by its legacy products, even those that are no longer supported. It’s clearly the right, and moral, thing to do.
I am fond of my Amazon Echo smart speaker (other brands are available). Some would say perhaps too fond. I am wont to ask, “Alexa, do I look good today?” and ‘she’ sometimes says I look great, which is in equal measure reassuring, but also rather unlikely.
By the way, before you ask, I do know Alexa is not really a ‘she’, but a combination of basic machine learning, speakers, a microphone and that mysterious Internet ‘thingy thing’. But I digress.
You see, a report this week by the BPI and Entertainment Retailers Association has asserted that the Amazon Echo and similar devices from companies that will never, ever be evil, has suggested that such smart speakers will buoy music streaming but have a negative effect on radio.
The logic is that said devices, with their ability to stream vast arrays of music at relatively low cost, will make it less likely that people who own such devices also, or ever, listen to the radio.
So Alexa, not video, could kill the radio star.
Smart speakers replacing other audio devices
An estimated 27 million smart speakers were sold last year worldwide and most people who have bought one listen to music on them. More to the point, the report found that 39% of smart speaker owners said that time they would have spent listening to the radio is now spent listening to music streamed to their smart speaker.
If that all sounds a little bleak for radio stations, may I offer a few words of comfort? Yes, it’s quite possible that smart speakers will temporarily distract listeners from their FM or DAB radio stations, as they become accustomed to being able to stream a huge amount of music to their speakers, on demand – indeed without even getting off the sofa.
But in my view, that may be a temporary aberration. In this era of fake news or ‘alt-news’ I know more and more people that are looking to some sort of curation of their news sources – an editor, a pundit, an organization with checks and balances – that they can trust more than Twitter, Facebook, or the myriad of social websites that we tend to visit every day.
There’s also the fact that my Echo will play most of the DAB radio stations for those times I want to listen to the news, some banter or unexpected songs.
Voice vs commodity music streaming
But the commoditisation of streaming music (Amazon says there are “tens of millions” of songs on Amazon Music and it’s about £1 a week to subscribe) may, in fact, make radio stations focusing on the spoken word the winners in all of this. BBC Radio 4’s figures have been weathering the storm, and chat radio LBC has seen some exciting growth on the back of challenging caller and celebrity or politician interviews.
Are smart speakers going to kill off radio? No, they aren’t. Might they force incumbent radio stations to consider their mix of music and spoken word? Probably, yes. In the mean time, have a listen to Frank Skinner on Absolute Radio on a Saturday morning. It’s irreverent and witty, but also have a listen to how the music – though frequent – takes a back-seat to the humour. In the face of Amazon Echo speakers and the like, the best DJs are fighting back with aplomb, and they are anything but traditional disc jockeys. Echo won’t kill the radio star, at least not in my lifetime.
But what do you think? Am I talking out of my Tweeter? Drop me a comment, I’d love to hear your views.
Amazon Web Services (AWS) announced its entry to the graph database market at its AWS reINVENT conference in Seattle in November last year. It was a notable announcement for a couple of reasons: it was the first graph database from the company (it offers a range of relational and NoSQL databases as a service). But it also shone a rather bright light on a database category that has often been considered niche, complex and expensive.
Neptune is currently in preview before it reaches general availability, but we expect that to happen soon. So should you be bothered?
A graph database is one that uses graph structures to enable the data to be queried, using the concepts of nodes, edges and properties to represent and store data. The key concept is the fact that the graph directly records the relationships between different data items in the database. Because the graph links related objects directly, it means those that have a relationship with one another can often be retrieved in one operation.
In relational databases, there are no such direct connections between related objects as data is stored in rows and columns. To create a relationship between different elements developers must write a ‘join’. But joins can become unwieldy and affect database performance.
The characteristics of graph databases enable the simple and fast retrieval of complex hierarchical structures that would be harder or even prohibitively time-consuming to model in relational databases.
The slight drawback with graph databases is that they cannot easily be queried with the de facto querying language for relational databases, Structured Query Language (SQL). Not only that, but in the graph database world there is not yet an equivalent de facto query language — there are a number of industry standard languages but there is likely to be a shakeout of some of these as graph databases become more popular and a clear winner possibly emerges.
Amazon says it built Neptune specifically for the cloud, which has its pluses and minuses. The drawback is there isn’t an on-premises version. The advantage though is that due to its economies of scale AWS tends to be able to offer good value subscriptions. As with other AWS managed services Amazon Neptune is highly available, with read replicas, point-in-time recovery, continuous backup to Amazon S3, and replication across AWS Availability Zones.
It can store billions of relationships and the graph can be queried with milliseconds latency. Neptune supports encryption at rest and in transit. As for that thorny issue of which query languages to support, AWS has hedged its bets with the option of Apache Tinkerpop Gremlin or SPARQL (Microsoft’s cloud graph offering, Azure CosmosDB, supports Gremlin or Gremlin-compatible languages such as Apache Spark GraphX).
I would have liked to see the addition to both of Cypher, a language developed by graph database pioneer Neo4j, as we believe it has very widespread adoption. Neo4j donated it to the openCypher Project in 2015 and as well as Neo4j it’s supported in SAP HANA Graph, Redis and AgensGraph databases.
Use cases and early adopters
Early adopters of Neptune are likely to be existing AWS users who have some or all of their data in the cloud already: AWS already offers a range of databases including relational and NoSQL options.
Amazon envisages that Neptune will power graph use cases such as recommendation engines, fraud detection, knowledge graphs, drug discovery, and network security. Security is probably the most common area where graph databases have been pressed into action, but they are also used in logistics, supply chain management, master data management, life sciences, e-commerce and even the hospitality industry.
Companies having a play with Neptune in preview include AstraZeneca, Thomson Reuters, Siemens, and the Financial Industry Regulatory Authority (FINRA). Amazon has been looking into how it can use it to improve its own Amazon Alexa system.
I believe AWS’ move into the graph database space is significant for the sector. It will make it simpler than ever for people to have a play with a graph database inexpensively. With Neptune, you don’t need to worry about hardware provisioning, software patching, setup, configuration, or backups.
It’s not that there are not other graph-as-a-service offerings, but few have quite the reach of AWS. With so many companies already having at least some of their data on AWS, this is an opportunity to see what a graph database can do for you.
There are too many graph databases to mention them all here, but here is a selection of firms large and small (in alphabetical order) to add to those mentioned above. Most offer some kind of pre-production free trial, so you can kick the tyres before you jump right in.
Do you have any experience of using graph databases? I’d be interested to hear your thoughts in the comments section.
IoT Back to Basics, chapter 4: IoT projects risk failure without careful consideration of data management processes and analytics. Their ultimate goal after all is to glean valuable information from data coming from the ‘things’ on the network – sensors and smart devices – in order to act on it.
So I thought I’d look at some of the novel trends in in-memory data processing: in-memory databases as well as data fabrics and data streaming engines.
The use of memory in computing is not new. But while memory is faster than disk by an order of magnitude, it is also an order of magnitude more expensive. That has for the most part left memory relegated to acting as a caching layer, while nearly all of the data is stored on disk. However in recent years, the cost of memory has been falling, making it possible to put far larger datasets in memory for data processing tasks, rather than use it simply as a cache.
It’s not just that it is now possible to store larger datasets in memory for rapid analytics, it is also that it is highly desirable. In the era of IoT, data often streams into the data centre or the cloud – the likes of sensor data from anything from a production line to an oilrig. The faster the organization is able to spot anomalies in that data, the better the quality of predictive maintenance. In-memory technologies are helping firms see those anomalies close to, or in, real-time. Certainly much faster than storing data in a disk-based database and having to move packets of data to a cache for analytics.
I expect take-up of in-memory data processing to accelerate dramatically, as companies come to grips with their data challenges and move beyond more traditional data analytics in the era of IoT. In-memory databases are 10 to 100 times faster than traditional databases, depending on the exact use case. When one considers that some IoT use cases involve the collection, processing and analysis of millions of events per second, you can see why in-memory becomes so much more appealing.
There’s another big advantage with in-memory databases. Traditionally, databases have been geared toward one of two main uses: handling transactions, or enabling rapid analysis of those transactions – analytics. The I/O limitations of disk-based databases meant that those handling transactions would slow down considerably when also being asked to return the results of data queries. That’s why data was often exported from the transactional database into another platform – a data warehouse – where it could more rapidly be analyzed without impacting the performance of the system.
Hybrid operational and analytical databases
With in-memory databases, it’s becoming increasingly common for both operational and analytic workloads to be able to run in memory rather than on disk. With an in-memory database, all (or nearly all) of the data is held in memory, making reads and writes an order of magnitude faster – so much so that both transactional duties and analytic queries can be handled by the same database.
There are a number of in-memory database players vying for what has become an even more lucrative market in the era of IoT. The largest incumbent database vendors such as Oracle, IBM and Microsoft have added in-memory capabilities to their time-tested databases. SAP has spent many millions of dollars educating the market about the benefits of its in-memory HANA database, saying it will drop support for all other third party databases under its enterprise software by 2025. There are also smaller vendors vying for market share such as Actian, Altibase, MemSQL and VoltDB.
Data grids & fabrics
Then there is the in-memory data grid (sometimes known as a data fabric) segment. This is an in-memory technology that you ‘slide’ between the applications and the database, thereby speeding up the applications by keeping frequently-used data in memory. It acts as a large in-memory cache, but using clustering techniques (hence being called an in-memory grid) it’s possible to store vast amounts of data on the grid.
In recent years their role has evolved beyond mere caching. They still speed up applications and reduce the load on the database, and have the advantage of requiring little or no rewriting of applications, or interference with the original database. But now as well as caching, they are being pressed into action as data platforms in their own right: they can be queried (very fast, in comparison with a database), they add another layer of high availability and fault tolerance – possibly across data centers – and they are increasingly being used as a destination for machine learning.
There are data grid offerings from a handful of vendors, amongst them Oracle, IBM, Software AG, Amazon Web Services, Pivotal, Red Hat, Tibco, GigaSpaces, Hazelcast, GridGain Systems and ScaleOut Software.
Data streaming engines
The third category, streaming, is also notable in the context of the Internet of Things. Data streaming involves the rapid ingestion and movement of data from one source to another data store. It employs in-memory techniques to give it the requisite speed. Streaming engines ingest data, potentially filter some of it, and also perform analytics on it. They can raise alerts, help to detect patterns, and start to form a level of understanding of what is actually going on with the data (and hence with the sensors, actuators or systems that are being monitored).
While streaming was largely confined to the lowest-latency environments, such as algorithmic trading in the financial sector, more and more use cases in the IoT space are latency sensitive: e-commerce, advertising, online gaming and gambling, sentiment analysis and more.
There are relatively few vendors with data streaming technology. But they include IBM with Streams, Amazon Web Services’ Kinesis in the cloud, Informatica with its Ultra Messaging Streaming Edition, SAS’ Event Stream Processing (ESP), Impetus Technologies with its StreamAnalytix and also TIBCO, Software AG and SAP (which bought StreamBase Systems, Apama and Aleri, respectively).
Smaller competitors include DataTorrent, which has a stream processing application that sits on a Hadoop cluster and can be used to analyze the data as it streams in, and SQL-based event-processing specialist SQLstream. Another young company is Striim.
In the open source space, Apache Spark Streaming and Apache Storm both offer streaming – most vendors have added support for Spark rather than Storm. But that, as they say, is a story for another day.
IoT Back to Basics, chapter 3: It’s no surprise that security and governance are important considerations when it comes to the IoT, but quite how incredibly important they are may not be immediately obvious.
Ensuring that users of IoT systems and smart devices remain safe and secure – which requires that their data stays protected and carefully governed – is vital if businesses and public sector institutions are to initiate successful IoT projects. There isn’t just the risk to a user’s privacy, and the possibility of big fines from regulatory bodies when things go awry, but also the issue of reputational risk and the commercial consequences of confidence in your brand being undermined.
Of course, security should be high on the agenda in all areas of IT. A targeted and sustained ransomware attack on the NHS, in May last year, was just one example of how sophisticated some of the hackers – and their malware – have become. At a machine data analytics conference last year, the chief security officer at Travis Perkins, a British builders’ merchant and home improvement retailer, told us that his organization had faced 3,851 ransomware attacks in just one month last summer.
The extra problem with IoT is that it vastly increases the potential ‘attack surface’ – there are more connected devices and gateways, and hence more areas of potential vulnerability, which gives those with nefarious intent greater opportunity to wreak havoc. And while many existing technologies and data governance methodologies can also be used in the era of IoT, they cannot make up for the broader attack surface.
Some of the ‘things’, such as sensors, are relatively dumb and therefore unlikely to bring much gratification to hackers. There’s not a huge amount of twisted satisfaction to be gained from interrupting temperature or wind-speed readings from a sensor in a wind turbine, for example.
But when you consider that IoT also includes the likes of connected vehicles, wear-at-home medical devices, industrial and hospital equipment, you can see why security is such a vital consideration.
For instance, in 2015 a group of researchers from the University of California, San Diego, discovered a serious weakness in vehicle security that allows hackers to take remote control of a car or lorry, thanks to small black dongles that are connected to the vehicles’ diagnostic ports.
These are common in both cars and lorries, fitted by insurance companies and fleet operators, as a way of tracking vehicles and collecting data such as fuel efficiency and the number of miles driven.
But the researchers found that the dongles could be hacked by sending them SMS text messages, which relayed commands to the car’s internal systems. The hack was demonstrated on a Corvette, where the researchers showed they were able to apply the brakes or even disable them (albeit as long as the car was at low speed).
You can imagine the repercussions of such a hack as we move ever-closer to driverless cars.
There have been other worrying security lapses around IoT that give pause for thought. In 2013, for instance, the US Federal Trade Commission (FTC) filed a complaint against TRENDNet, a Californian maker of home-security cameras that can be monitored over the Internet, for failing to implement sufficient security measures.
TRENDNet’s cameras were hacked via the Internet, leading to the display of private areas of users’ homes on the Web, and allowing unauthorized surveillance of adults as well as children going about their usual daily lives. As well as an invasion of privacy, there was the potential that such covert surveillance could be used to monitor the comings and goings of the occupants of a premises, and hence give rise to further criminal activity once the hacker knows when there is no one at home.
Clearly, some IoT initiatives have different risk profiles to others. For instance, ‘white hat’ hackers last year demonstrated that they had been able to hack into a smart domestic appliance network and turn off ovens made by the British company AGA. Being able to turn them on and adjust the temperature would be more dangerous, but the ramifications are still worrying.
Another penetration testing company discovered that hackers could remotely compromise a connected kettle with relative ease and thus potentially gain unfettered access to a person’s wireless network, from which they could change DNS settings and monitor all web traffic for access to bank accounts and other sensitive data.
It’s obvious that the companies involved in implementing IoT need to be just as sophisticated about their security processes and protocols as the most sophisticated hackers – but time and again we have seen companies outsmarted by either ‘white hat’ or, worse, ‘black hat’ hackers.
The potential security risks around IoT are very real
Organizations contemplating the benefits IoT projects (or in the case of local or federal government, their citizens) would be wise to consider security and data governance very carefully indeed. Authentication and authorization technologies are likely to be necessary. Data masking (removing attributes that would enable a hacker to identify specific people and their habits, for instance) may also be called for, and in some cases even mandated by law.
Ensuring privacy is also an issue. While some consumers or citizens are quite happy to share various data with organizations, others are not. Organizations must therefore ensure that they ask users to ‘opt in’ to IoT-related projects or systems, rather than opting them in without explicit consent (even if they subsequently offer an opt-out).
Companies that don’t do this run the risk of annoying customers and falling foul of auditors and legislators. If potential fines are not sufficient to deter some companies from taking security and data governance seriously, the potential reputational damage certainly should be!
It’s all too easy to conflate the kind of AI being hyped in the industry at the moment with the science fiction notion of machine sentience. We are still a long way from the latter, though, whether you see it as WALL-E or the Terminator.
What’s mostly on offer today from IT vendors and service providers is really just advanced data analytics. With the power and scalability of modern cloud platforms, rules and inferences can be constantly updated and refined as more data is accumulated (machine learning), and applied in near real-time. This can of course create the illusion of intelligence, but sentient computers it isn’t – not yet, at least.
Having said that, we shouldn’t underestimate the potential for today’s kind of AI to have a huge impact on the way some things are done in business. If you work in IT, this is something you need to get to grips with sooner rather than later. Why? Because AI capability is rapidly becoming a lot more accessible, and before long will be pervasive across our application and service estates.
The pace at which things are evolving became clear to me over the course of a number of briefings and conversations I had towards the end of last year. This began with a session at IBM, during which a case study at major oil and gas company was discussed. The Watson ‘cognitive computing’ platform had been used to create a virtual assistant that was transforming IT support by providing users with advice and guidance via a multi-lingual text and speech interface. The results achieved in terms of service level metrics were impressive, but to get there required a substantial professional services engagement – i.e. lots of consulting time and expertise.
In contrast, I then had quite a different conversation with Salesforce.com, which has been acquiring, building and integrating AI capability into its cloud platform for a number of years. In the words of John Taschek, Senior VP of Strategy at the company, “A lot of what we are doing is aimed at making AI a seamless and embedded part of the business process”.
Moving AI into the software stack
One of the examples we discussed was advanced sales forecasting powered by Einstein – the overarching brand name for most things AI in Salesforce.com. The key point here is the notion that you shouldn’t need lots of specialist expertise or coding and integration effort to exploit the potential of AI. It will increasingly be a case of ‘switch on, configure and go’.
More recently, at its January Tech Summit in Birmingham, I heard Microsoft do a pretty good job of spelling out the different routes to AI goodness. If you have the expertise and want to get really ‘down and dirty’, the Azure platform is increasingly going to offer fine-grain AI and machine learning capability, right down to FPGA level. For mainstream developers who need to AI-enable their application without having to worry about the detail, higher-level services are offered so you can access natural-language functionality. For example, a set of APIs can hide all of the underlying AI complexity. Then, further up the software stack, we’ll increasingly be seeing AI smarts embedded seamlessly into Microsoft applications and tools, from Office 365 to its CRM and ERP offerings.
I’ve only mentioned three players here, but technology companies large and small, from Google and Apple to highly-innovative specialist vendors, will be surfacing AI capability in all kinds of different ways. That includes embedding it in the systems and security management tooling used by IT teams.
The upshot is that AI will increasingly find its way into the world of IT professionals – there really won’t be a way of avoiding it. So you need to starting thinking now about the implications in relation to changing user expectations, application design and implementation, service management and support, and not least, security, privacy and compliance.
When we surveyed several hundred IT professionals on the topic of All-Flash Arrays, one thing that came out was just how broad was the chasm in thinking between those whose organisations already owned and used AFAs, and those who did not.
Most current AFA users were positive about the technology’s value, both to the wider business and to IT specifically. However, non-users were much more likely to be cautious or even sceptical about the strategic value and operational benefits of AFA.
We also found that these two groups had quite different ideas of which business workloads work well on AFAs. For those with no direct experience, the top target workloads were database applications and virtual servers, both of which were thoroughly hyped up in the early days of AFA, of course.
Once again, familiarity with the technology had a robust effect: the experienced group were using AFAs to support a much broader range of workloads. As well as databases and VMs, they included online transaction processing, mobile apps and services, virtual desktops, big data, and real-time analytics.
Going beyond simple workload suitability, we also asked about using AFAs to enable business and IT transformation. Here, we were thinking about those changes that derive from Flash working differently from disk, such as its ability to deliver consistent performance and reliable quality of service. The majority of those with direct experience agreed that AFAs were a strategic enabler for both business and IT transformation, while those without direct experience were rather more cautious.
We were also thinking about the way AFAs bring more opportunities for automation, and sure enough the second most significant benefit reported in our survey was that they need less management and tuning. As well as the opportunity to free up skills and redeploy them to create real business value, this also implies less downtime resulting from ‘human errors’.
Of course, we didn’t know exactly why the non-users were non-users. It could be they were indeed sceptical of AFA’s value, or perhaps they simply couldn’t get the budget, hadn’t had a trigger to change, or thought that their applications weren’t appropriate for Flash storage. The result was the same though – actual experience is key to understanding the possibilities of the technology, and they didn’t have that experience, hence the awareness chasm.
There may also be an element of ignorance and working on outdated information. Not everyone is aware of how fast AFA technology has evolved over the last couple of years from the niche-oriented first generation systems, or of how quickly its effective price per GB has fallen. As a result, there is still some residual uncertainty and doubt about the enterprise relevance of Flash – doubt which our experienced users tell us is largely unwarranted today.
Either way, our research shows that, when it comes to understanding and achieving the potential of AFA, experience is a massive help. Once you have worked with it, you ‘get’ it.
But as the saying goes, there’s a first time for everything, and even if you don’t have direct experience to help you, you can still bridge that chasm and build a good business case. Reading our report (it’s free to download) will help when it comes to understanding just how many of your applications could benefit, for instance, as will talking to those who have already gone along the AFA route.
Then it’s careful planning, of course. Put that business case together, profile and test your apps – that’s a key tip from our experienced users – and make sure you choose a supplier with good post-sales support and the ability to advise on best practices.
And if you’re trying to sell the idea of investing in AFA to someone else in your organisation, remember that they might well have a rather distorted idea of what it’s good for!
Microsoft’s revenues are up, but compared to its biggest competitors — Amazon, Apple, Google and Facebook — its mindshare and perceived market relevance are down.
Satya Nadella’s Microsoft is still going through the process (and pain) of reinvention, morphing from the PC-based Windows & Office-centric company that everyone knows, to the cloud-based Azure & AI services company that investors want to have in their portfolios. But with so much riding on this phase of the company’s evolution, Microsoft must convince businesses, enterprises, governments, consumers and partners that it has something useful, if not essential, to offer across a complex mix of markets and sectors.
Normal people don’t think about ‘computing’
We’ll never really know what people thought about Apple’s recent ‘What’s a computer’ iPad Pro promo video, because Apple (unlike Microsoft) doesn’t let people add comments to its YouTube content. However, from a consumer perspective, the ad makes a valid point, in that no one playing a game, using a business application or doing their homework on a PC, mobile phone or tablet device ever thinks of this as ‘computing’.
The visible components of Microsoft’s ‘more personal computing’ strategy currently revolve around Windows 10 and the company’s range of adaptable, yet expensive, Surface devices. This combination will maintain the company’s relevance in its traditional desktop domain, but from a platform perspective, Microsoft needs to engage with the growing number of non-Windows, non-PC users.
A cloud that can listen, learn and predict
For normal people, the word ‘computing’ is something that happens in the cloud or in that place we call the data centre. Of course, Microsoft has a major presence in both locations, and is therefore well placed to service the hybrid IT needs of organisations, but to remain relevant in the noisy consumer market, it must find new and authentic ways to convey the notion of it being useful or, better still, essential.
Microsoft failed in its attempts to make its Windows Phone platform either useful or essential during the mobile technology wave, so it needs think about the lessons it learnt as we enter the ‘digital assistants’ technology wave, driven by voices, smart speakers and intelligent devices. There’s a lot of hype around this topic, but very little of it features Microsoft’s own digital personality, Cortana. So, Microsoft must find a way to insert its intelligent cloud and intelligent edge technologies into the equation, and getting developers to build intelligent applications on its Azure platform is the most obvious option.
Nothing important happens in the office
Microsoft isn’t alone when it talks about reinventing productivity and business processes, but it’s one of the loudest. Office 365 is undoubtedly changing the way that productivity tools and associated services are delivered to employees and end users, but there’s scant evidence to suggest that it’s radically changing the way that people use Microsoft Office.
Organisations don’t really differentiate themselves by the way they manage file servers, email servers or telephony systems. OK, some users are a whizz when it comes to using Word, Excel and SharePoint, but it’s what happens outside of Microsoft Office that ultimately matters to the organisation. This is why partners will ultimately determine the relevance of Microsoft in 2018, and why Microsoft needs them now more than ever.
This article is part of a series on the challenges facing major technology firms in 2018. For more, please see the main Write Side Up blog page.
IoT Back to Basics, chapter 2: In the era of the Internet of Things (IoT) it is becoming increasingly important to be able to process, filter and analyse data close to where it is created, so it can be acted on remotely, rather than having to bring it back to a data-centre or the cloud for filtering and analysis.
The other reason to implement analytics at the edge of the network is because use cases for IoT continue to grow, and in many situations, the volume of data generated at the edge requires bandwidth levels – as well as computing power – that overwhelm the available resources. So it’s possible that streams of data from smart devices, sensors and the like could swamp datacentres designed for more traditional enterprise scale needs.
For example, a temperature reading from a wind turbine motor’s sensor, that falls within the normal range, shouldn’t necessarily be stored every second, as the data volume can soon add up. Rather, it is the readings that fall outside of a normal range or signify a trend – perhaps pointing towards an imminent failure of a component – that should create an alert, and possibly be stored centrally only after that first anomaly, for subsequent analysis.
There are too many vendors in this space to produce an exhaustive list here. But it’s perhaps notable that last year, a company formerly known as JustOne Database performed a root and branch rebranding exercise. It renamed not only its products, but also its company name, which is now Edge Intelligence. It told me it was seeing such good traction for its database – that can run on relatively compact servers at the edge of the network, a data-centre or the cloud – that it changed its name after over six years in the business.
So what are some of the characteristics of edge analytics that you might want to consider if you are trying to push at least some analytics to the edge?
Standards and protocol translation
Although there is likely to be a shakeout of some of the standards in this space, opting for technologies that support standards is likely to make future integrations easier. Again there is a vast array of standards and API’s in this area. Standards and protocols include POSIX and HDFS API’s for file access, SQL for querying, a Kafka API for event streams, and HBase and perhaps an OJAI (Open JSON Application Interface) API to help with compatibility with NoSQL databases. There’s also the need to be able to support older, proprietary telemetry protocols so that legacy equipment (that often have lifetimes measured in decades) can been connected to more modern IoT frameworks. This is especially true in the industrial space, where IoT is of particular value for the likes of predictive maintenance.
Distributed data aggregation
This is to some extent the bread and butter of edge analytics, providing high-speed local processing, which is especially useful for location-restricted or sensitive data such as personally identifiable information (PII), and can be used also to consolidate IoT data from edge sites.
This refers to technologies that adjust throughput from the edge to the cloud and/or data centre, even with occasionally-connected sensors or devices.
Combines operational decision-making with real-time analysis of data at the edge.
Security and identity management
End-to-end IoT security provides authentication, authorization, and access control from the edge to the central clusters. In certain circumstances it will be desirable to offer secure encryption on the wire for data communicated between the edge and the main data centre. Identity management is also a thorny issue: it’s necessary to be able to manage the ’things’ in terms of their authentication, authorization and privileges within or across system and enterprise boundaries.
Delivers a reliable computing environment to handle multiple hardware failures that can occur in remote, isolated deployments.
Integration with the cloud
Even if not now, there may be a requirement in the future to have good integration between an edge analytics node and the cloud. This is so that alert data and even ‘baseline’ data points can be stored in the cloud rather than in one’s own data centre. In this regard integration with your cloud provider of choice – if you have one – would be a wise idea. If you don’t already do much in the way of data processing and storage in the cloud, some of the likely execution venues in your future could include Amazon Web Services, Google Cloud Platform or Microsoft Azure, but it wouldn’t do any harm to know there is support for the open source OpenStack infrastructure as a service (IaaS).
Edge analytics has come on leaps and bounds in the past several years as IoT use cases have shaken out. At the very least it might be worth asking if edge computing has a role to play in any IoT projects that you may be thinking of embarking on.