Category Archives: Enterprise Architecture

Microservices Architecture – oh no, not another one?

Haven’t blogged here for a while ‘cos nothing much has captured my attention…but when something does capture your attention…

As mine has been by a tweet from the Open Group on Microservices Architecture, linking to a piece by Eric Knorr on Infoworld and through to another by Martin Fowler. I may be ‘behind the curve’ (gah!) here but this seems to be some interesting stuff.

Martin starts his piece with ‘”Microservices” – yet another new term on the crowded streets of software architecture’ – hence oh no not another one to add to the business, enterprise, technical, service-oriented etc architectures that are crowding the landscape.

But this seems to be some interesting stuff.

You’re best left to read the articles for yourself as they put it much better than I can, but just a few teasers from/prompted by Martin.

  • Componentisation via services – not a new idea, but not one that’s been fully realised either – even at the business architecture level, we want components! & not having to keep on reiterating those massive, monolithic, ERP-style software renewals…
  • Organised around business capabilities.
  • Products not projects – ‘You build , you run it’ (Amazon). 
  • Decentralised governance – use the best tools for the job. If they don’t want to dig a hole, don’t make them use a spade!

Well, got me thinking anyway…

Digging a hole with Enterprise Architecture #entarchs

Some Sunday morning rambling & a couple of old Dilbert cartoons which may be pertinent & should at least be a amusing…

dilbert2

So, you know the scenario. You’re in there with the senior management team & they want an answer to the question: why should we invest in this thing called Enterprise Architecture?

& the answer is: I dunno; or, you tell me; or maybe, you shouldn’t.

Because they are, of course, asking the wrong question. They should be saying: how do we improve our customer experience? Or, how do we reduce the cost of procurement? Or, how do we decide what we could target as shared services? Or…a thousand other critical business questions. To any or all of which EA may be part of the answer.

But you’ve been invited along to a meeting to explain why they need EA. So you’ve come to the wrong meeting.

& because they’re asking the wrong question, you’ll probably have to give them the wrong answer.

& you can’t directly tell them they’re asking the wrong question. Well, go for it if you feel like it. After all, they’re only senior management…

There’s a lot of discussion in EA circles about how exactly do we sell EA to senior management? & the answer is, of course, you don’t.

You don’t sell someone a spade if they don’t know they need to dig a hole. At least, I hope not – don’t want to be in the snake oil business. So senior management need to know what problems EA will solve, or what new opportunities EA can open up – they need to see what bigger better hole they can dig (themselves into?) with this shiny new implement (maybe not the best analogy, then, but you have to start somewhere).

So EA isn’t always the answer – it depends on the question, & you have to make sure the right questions are being asked then you can make sure you have the right answers.

So in that meeting, you need to get them to ask the right questions – ‘I dunno’ may be a pretty good first answer – followed by, you tell me what you need to do, & then I’ll tell you if EA can help you do it. Or, maybe not…because…

If you’re an Enterprise Architect, you might think EA is always the answer, or at least part of it – & I agree, you’re probably right. But the best way to test a hypothesis is to assume that you’re wrong – & maybe the customer doesn’t really need to dig a hole, at least, not until they’ve chopped down that tree. Great thing about EA is, it’s more like a Swiss Army knife than an axe or a spade – but if the customer only wants to dig that hole or chop down that tree, some kind of multitool might seem like overkill. So, find that hole that needs digging, use EA to help it get dug, & then…

dilbert1

UCISA Enterprise Architecture Community of Practice Start Up Meeting

2nd May 2013: a group convened at Liverpool John Moores University to discuss the setting up of a UCISA Enterprise Architecture Community of Practice (EACP).

Following work by Luke Taylor & the UCISA-CISG, & commissioned by the UCISA Executive, the objective of the meeting: to baseline where we were up to with EA, and work out how this Community of Practice thing was going to operate, as it’s a bit of a new departure for UCISA & those previously involved in the JISC EA Practice Group. Representatives from 8 institutions, JISC InfoNet & the Open Group, and facilitated by Richard Chapman, of Richard Chapman Consulting.

This is a picture of where we started:

EA0213

I’m not going to do a blow by blow account of the day, as that will come out in due course, but just wanted to highlight a few things that were key or of interest for me.

  • Baselining EA. Various people were not surprisingly at various stages, although there were common themes: how to keep up momentum, how to convince the business, how to get resources. Participants were asked to come up with ‘twEAts’ (140 characters about EA on a postit – thank you David Rose of the Open Group for that one, bit close to our great Prime Minister’s comments about what too many tweets make…). These got quite zen-like in parts – particularly this offering from Patrick O’Reilly/Bolton: ‘the curate has a fresh egg, but no eggcup, spoon and is waiting for a chance to eat’. I thought we might have invented Zen EA – but then discovered this: ‘Zen and The Art of Enterprise Architecture (Open Group Conference Newport Beach 2013)’ – although I may be being a little unfair, having not viewed it in detail, at 83 slides, some of dazzling complexity, I’m not sure this is in the spirit of Zen, or even EA…
  • Building a Community of Practice. Lots of interesting discussion on this, which will be rolled up in a wiki & considered further – but for me, the main issue was around leadership. The key objective of the EACP is to become self-sustaining – to not need a leader, or a Chair, or single person or even small group who keep it all going – leading to the usual leaders/followers scenario. Seems like a kind of variation on situational leadership – the idea that there is no one best leadership style, but that leadership is adaptive to the situation – which in some situations may mean following. So perhaps we have situational leadership & followership as well. For the EACP I would envisage a scenario where individuals would either lead or follow depending on the situation – which I think is what I’ve just said. Or to sum up, ‘We are all leaders now’.

Which is about all for the moment – the EACP is a work in progress & progress on this first day was excellent, we have actions, ideas for how to work, deadlines & all manner of good stuff, including the all important hash tag (#UCISA_EA) – so thanks to all the participants & also Richard Chapman for his great facilitation, kept us almost both on time & under control!

Also worth a look at Dave Berry from Edinburgh’s post on the same event.

& here’s a picture of where we finished:

EA2013b

UCISA Cloud Computing Review – 31/10/12 Nottingham #uccr

Usual random notes that may be of interest…see Programme for access to presentations.

Simon Marsden/Edinburgh: Introduction

What’s in/what is a cloud? Standard definition. Gartner view: high security, privacy or competitive advantage – otherwise outsourced or offsite. But myths abound:

  • Costs will be predictable – not to start with!
  • Costs will be lower – no!
  • More resilient – if you make it so
  • Disaster recovery – as above
  • Assured performance – or noisy neighbours?
  • Don’t know where data is? – probably do

User perspective: I get what I want; it just works.

CIO challenges: not too many – scale, exit strategy, data loss, how to charge, integration. & need to be seen as part of the solution, not an obstruction to be bypassed.

Things to look at: Coursera, learning platform for MOOCS; Office 365; Unidesk shared helpdesk; elastic cloud; trust; app store; reliability; green.

Agility: key issue. Running MOOCS with Coursera for over 100k students, quick deployment. Important to be there at the start, couldn’t have done it themselves.

Technology Enhanced Learning cloud – hybrid of supported through to non-supported, but provided within app store – flexible way of choosing tools.

Resilience – Site24x7, full scale monitoring.

So Cloud enables: swift innovation, reduced costs, greater choice for users. Not an option: can be difficult to understand, needs leadership.

Great start!

Richard Maccabee/ULCC: Delivering cloud services to the sector: ULCC’s experience of the range of options

Is it good use of public money for every HEI to have its own data centre? (or two…) – probably not, but is it public money anymore? All about services to paying customers, perhaps, perhaps sadly.

Approach: OS; shared services; vendor partnerships; evolutionary; hybrid: supplier, based in the community.

Service portfolio includes middleware & applications: ESB; Digitary (DARE!); Agresso Finance System, for about 10 institutions.

About 300 customers, across the sector & significant number of FE.

Issue: single data centre, in London – not good choice. So second site commissioned for 12/13, & looking at partner for Cloud IaaS. ULCC data centre will become point of presence for the Cloud. First bid submitted for provision of full IaaS for FE college.

MMU: very good student-focused VLE built around Moodle but including distributed services.

Drivers: mobile; BYOD/DIY; increased customer expectations; SOA ref Nexus ESB/DARE/Digitary, important for SaaS. Looking into Student Systems. <oh right, hosted then?>

Good that SOA – & to my mind hence EA – still seen as key component.

Ed Carter/Leeds: YHMAN – A Data Centre Without Walls!

Not Ed’s fault, but have a feeling I may have heard this before!

8 Universities collaborating in shared virtual data centres: enable economies of scale, balance asset utilisation, meet carbon reduction commitments, enable growth, enhance service standards, adhere to stringent security arrangements. Built on JANET/YHMAN network. Not that far away! & JANET6 increases potential reach/scalability. Storage for research, big data; HPC capability.

So essentially, promise of Cloud = ‘EaaS’: Everything as a Service. The Virtual University – the ‘University In A Box’?

Pleasantly surprised, if I’ve understood this right, by transition of technical/network staff into service support. Backed up by supplier support – HP.

It works!

Shan Rahulan/janet: Janet Brokerage

Set up to: promote cloud; reduce hurdles to adoption; create efficiencies. Like YHMAN, applying network approaches to data centres/services.

Usual drivers as previously referenced; also noted that adoption/governance is formative, & integrating clouds not straightforward. Customers are in the Cloud space, institutions/IT depts aren’t (entirely, yet).

Cost savings: not proven; current costs not understood, so savings hard to define. Hard to compare like with like.

Working with BUFDG, UCISA, JISC & UUK on modelling costs – which should be helpful!

Also looking at Amazon & Dropbox to establish a sector offer.

Olly Butters/Leicester: Building BRISSkit on the Eduserv Cloud

BRISSkit – Biomedical Research Infrastructure Software Service kit. Funded by UMF Cloud initiative to deliver in the Eduserv Cloud.

Good practical example of delivery of a service in the Cloud.

Cloud challenges – like ‘Hard to give up hardware – I can’t go and kick it’. Interesting thought – can you kick a cloud?

Working through governance issues/toolkits.

Mini Panel Session: Shan Rahulan; Ed Carter; Olly Butters: Matt Johnson/Eduserv

What are the barriers? People still want their servers on their site (‘Can you kick it?’ ref A Tribe Called Quest, perhaps). Again, internal services not fully costed/full costs not understood.

Simon Marsden: you need a burning platform – agree entirely. For Edinburgh, this is agility. Must be a real business/service driver because otherwise won’t want to meet the costs of change – cost saving not enough.


Stephen Booth/Coventry: A Managed Storage Service on a Hybrid Cloud

Can no longer do ‘just an IT project’ – I thought it was quite a while since you ever could? Business not IT initiative. Business drivers: cost reduction; flexibility; enhanced support; DR. Importance of supplier/partner relationship – joint customer/partner management of project & service delivery – partner will not let it fail: so have to make sure they have some skin in the game.

John Waters/University of West London: The North Face of Office 365

The University formerly known as TVU. 12 years on the HEFCE at-risk register. Rebranded as UWL April 2011. Bit of a mess of overlapping/inconsistent systems for student services.

Wanted to move to functional Cloud-based Student Portal & facilities, rich use, back end integration, SSO, IDM etc.

Approach: generated SMT buy-in through Microsoft-led activities at TVP. Project team. Focus groups. Piloting & proofs of concept. Got some help! (Fulcrum). 3 parallel work streams: infrastructure; IDM & data migration; portal design & back-end integration, Agile/Scrum, .net into O365 Sharepoint.

How does it work? Create accounts in Cloud (SIS = Capita Unit-e) along with authentication. FIM & ADFS. Systems Integration – .net to Sharepoint. Some use of Lync.

What’s in it? See presentation.

Content & administration issues ref content from Schools. Feedback from students – overlap with groups in Facebook. Systems integration: interestingly, no harder than doing on-site.

We did it (within 5 weeks) & students like it. Say no more. & next Phase 2 – mobile, Lync, staff email, SSO for VLE & SIS, distance learning, etc. Phase 3 – August 2013 – further data integration, social features, business portals.

Fulcrum Worldwide for site customisations & systems integration.

Useful day – & now, publish & be damned, & out of here to catch that slow train to Southport…

#UCISA12 UCISA Management Conference 2012 Last Day, Last Post

Session1 – A cloud by any other name… Simon Daykin/CTO, Logicalis

Are we ready for cloud? From a technology perspective, yes. Loughborough – totally cloud. Transition to customer centric experience. janet brokerage. Really excited. Put your arms around that experience. Perfect storm. Immersive. iPad6. No longer central service that you can put your arms around. Lot of putting arms around things. Current approach is no longer viable way of delivering IT. Is this telling me anything? Can’t control the endpoint – so don’t try…liberate our services. Put all our services behind the firewall & then…put our arms around them! Should have seen that one coming. The general internet forms a core part of what we do! Noooo! Really? *fake surprised expression* Private cloud – that you need to wrap your arms around! Going forward. What I would say is. You can’t put your arms around a memory, can you put them around a cloud? Wrap them in a security blanket. Mutual authentication & posture. Collaborative & connected. Cool. Know what to do, just need to do it.

IT services in the HE sector: a private sector perspective, James Salmon/CIO, BPP

Global requirements for efficient, fast, standardised IS. Don’t start with technology – start with process! Actually, don’t start with process – start with people. Have to engage people first. People, process & technology – here we go again. We know the ducks, but can we get them in a row?

Review the data, understand the data, be the data…& then be out of here.

 

#UCISA12 UCISA Management Conference 2012 Day 1 Part 1

14th March 2012. At the Celtic Manor Resort Hotel, somewhere near Newport. Wondering whether or not to tackle the assault course outside my window. Maybe not. & then there’s the lobby…abseil, perhaps?

Session1: Welcome, Peter Noyes/Vice Chancellor, University of Wales, Newport

Working with business to create the Entrepreneurial University, with the best VLE that money can buy – that’ll have to be a new one then…clearly recognises the strategic importance of ICT, which is a good start.

Session2: Providing sustainable public sector ICT services for Wales, Gwyn Thomas/CIO for Wales

From technology to information…

Lot of commonality between HE & the rest of the public sector, as indcated by the content of the exhibition – but need to identify where collaboration adds value. Strategy of critical mass, not consensus – not everyone needs to agree with everything. Cloud: the critical mass is nearly there.

Building the social architecture on which the technical architecture will rest – sounds good. Devolved/collective leadership – don’t know what goes on in a scrum, but can’t take any part of it away. CIO Council for Wales – strategy. Delivery – regional. Public sector design authority – interesting idea – experts responsible for standards & architecture. Wonder how much authority it really has? Once for Wales is the default option. Social architecture – getting people together to work in common. Technical architecture flows from social. Banned the word pilot – had more pilots than the RAF – now seen as first steps, if don’t work, stop.

Technical governance – national architecture. Programme governance – regionally delivered. Benefits – locally realised. Sounds about right.

Our Age of Anxiety is, in great part, the result of trying to do today’s job with yesterday’s tools – with yesterday’s concepts” – Marshall McLuhan.

Moving away from the institution to the person – person-based services.

Too many data centres – value of baselining, measuring – 24 with excess capacity, 19 with insufficient, must be better way of doing things.

Defining the approach for CEOs: 4 themes – buying better, using smarter, working together, working differently. & that’s the strategy. Discussion: ICT is just the enabler, what is it you want to enable?

Moving from lots of infrastructure & not many online services, to vice-versa. Balancing economies of scale with secure access. Base security on the nature of the services required, not vice-versa.

Everything is marching inexorably towards the cloud, on the cusp of a new kind of ICT department, little infrastructure, specialist apps, intelligent customer, technical governance & design, collaborative working, information not technology.

From systems to services! Well, whaddaya know!

To be valued, information must be: well designed & presented, trusted & authoritative, benchmarked against the best, rated by users, & worth something…

 

 

 

 


 

JISC/LFHE Strategic ICT Toolkit Closedown Meeting

To the University of Nottingham for the final meeting for the project team – JISC/LFHE/JISCInfonet – & institutions who have participated in the SICT field test projects. Taxi ride from Nottingham station taking in a firebombed police station – not something you see everyday. Possibly a more than usually garbled commentary, but it was a very worthwhile initiative so I feel the need to get something down.

First session of the day was for each institution to do a Pecha Kucha presentation on their project. The following are a few interesting bits & pieces I picked up…

Bruce Levitan/Manchester Metropolitan University raised the interesting issue of the impenetrability of Enterprise Architecture for those from a non-technical background. Good point, although my experience suggests that as usually presented those from a technical background don’t necessarily find it any easier. Also see my closing comments. Anyroad, he suggested that some more engaging way of presenting EA was needed, & showed part of what looked like a very good Gartner EA animation. Which took me back to illustrating SOA/EA to LJMU management using the JISC SOA animation, & discussions in the EA Practice Group & beyond about how great it would be to have something for EA. Well, it still would.

Ian Hall/York had approached the project through the interesting mechanism of 1:1 interviews with senior managers using the toolkit. Comments: again, EA could be off putting, & not necessarily necessary/essential? Outcome – look into it. Equally, Shared services seen as a means to an end rather than necessarily a dimension on their own. Highlighted need to work on communications! & Information Governance. Started thought process. Will run again.

City of Glasgow College. Difficult to get off the ground but got conversations going. Separate toolkit for FE? EA awareness only with technical staff.

Middlesex. Academic view. Challenged dimensions ie not the ones they thought were relevant eg discarded shared services. Communications! Toolkit – jargon; limitations in spreadsheet; duplication, poor flow of questions. First real attempt at measuring ICT provision. Need to collect data over time – longitudinal.

& then things become rather more garbled than usual as it was my go & I suffered Powerpoint meltdown. My attempt at setting up a timed presentation for my Pecha Kucha (I can’t help thinking of gutta percha…) had resulted in an 0.06 second transition time between slides, which even my fast talking couldn’t keep up with. Attempts to fix it while continuing the presentation failed miserably, so I had to take time out – hence my recollection of other Pecha Kuchas is limited as I was labouring over a hot laptop, thanks for the loan, Lesley Huxley.

Clearly my new ‘zero or one’ strategy for slides is right for me – should never have gone for gutta percha.

So all I can say of the rest is: Coventry – Academic leaning, challenged the premises. Loughborough – scored Strategic! & devised their own simplified version. Bolton used it to support bringing in a Governance structure. & if I’ve missed you out completely whilst I was floundering around, apologies…

Next session: breakout to brainstorm around our projects. Lot of interesting stuff – particularly discussion around whether or not Shared Services should be a dimension, & the tension between generic/persistent dimensions – leadership, communications – & dimensions that may be time-limited eg if you did the exercise 5 years ago, or 5 years in the future, would EA & Shared Services appear? Emphasised the importance of focusing on the objective not the means ie an institution wants effective/efficient services; Shared Services may present a way of achieving these; an institution needs to understand the relationships between it’s people, processes & systems; EA is a way of articulating this. If you don’t go with Shared Services or EA it doesn’t mean you don’t need efficient/effective services or to understand relationships.

Also interesting discussion about how projects get prioritised/get senior managemnt backing, & where innovation fits into this. General rule:

Followed by a video conference session with Alex Hawker talking about the new Strategic Information Practice Group initiative, which will take on & progress the work of the EAPG and SICT, & the JISC call 11/11 for Transformation projects which will engage with these & other JISC-supported tools to deliver organisational change – very useful.

& lastly that I’m going to comment on, Lesley Huxley from the LFHE gave a rundown of how they might incorporate the SICT in their work, & led a very useful discussion around how it could be used in management development. One very straightforward idea: Senior management teams should be knowledgeable about the capability of ICT & what it can do for their business. They really should. Seriously. They should all be type 42s. They bang on all the time about the need to work smarter not harder, to be online, to engage with the digital natives – so plainly at the capability level they need to know about this stuff. So use the ICT Toolkit to find out if they do…

Also interesting discussion about how projects get prioritised/get senior managemnt backing, & where innovation fits into this. General rule: cheaper & better; can also go with better & the same; cheaper & the same; & sometimes even cheaper & worse, if a service is necessary but only needs to work rather than add value. Then raises the question: what about innovation? New things that by definition can’t be cheaper or better because they don’t exist. Suggested that innovation could be another SICT dimension.

Final comments:

– there was a lot of discussion around whether EA & Shared Services were really important strategic dimensions. I think the problem’s with the language. Using these terms can ghettoise them – they become magical technical things that can be put over there with the magical technical people – if you talk about services, which might be shared, or understanding relationships, which can be done with EA, you’re in a different ball game with a chance of winning.
– there still seems to be a lot of obfuscation around EA. EA is like Project Management – it’s just organised common sense. It’s not that complicated; it’s not as complicated as the experts make it seem; bring on the animation!
– Hell yes, JFDI