What if the OSS solution lies in its connections?

Imagine for a moment that you’re sitting in front of a pristine chess board, awaiting the opportunity to make your first move. All of the pieces have been exquisitely carved from stone, polished to a sheen. The rules of the game have been established for centuries, so you know exactly which piece is able to move in which sequences. Time to make the opening move.

You’ve studied the games of the masters who have preceded you and have planned your opening gambit, the procession of moves that will hopefully take you into a match-winning position. Due to your skills with modern automations, you’ve connected some of the chess pieces with delicate strings to implement your opening gambit with precision.

Unfortunately, after the first few moves, your strings are starting to pull the pieces out of position. Your opponent has countered well and you’re having to modify your initial plans. You introduce some additional pulleys and springs to help retain the rightful position of your pieces on the board and cope with unexpected changes in strategy. The automations are becoming ever more complex, taking more time to plan and implement than the actual next move.

The board is starting to devolve into unmanageable chaos.

Does this sound like the analogy of a modern OSS? It’s what I refer to as the chessboard analogy.

We’ve been at this OSS game for long enough to already have an understanding of all of the main pieces. TM Forum’s TAM provides this definition as a useful guide. The pieces are modular, elegant and quite well understood by its many players. The rules of the game haven’t really changed much. The main use cases of an OSS from decades ago (ie assure, fulfil, plan, build, etc) probably don’t differ significantly from those of today. This
“should” set the foundations for interchangeability of applications.

We see programs of work like ONAP, where millions of lines of code are being developed to re-write the rules of the game. I’m a big advocate of many of the principles of ONAP, but I’m still not sure that such a massive re-write is what’s needed.

It’s not so much in the components of our OSS as in the connections between them where things tend to go awry.

The foundation of all brilliance is seeing connections when no one else does.”
Richard Parkinson
.

This article distills ONAP from its answers back to the core questions. What if instead of seeking an entirely-new architectural stack, we focused on solving the core questions and the chessboard problem – the problem of connections?

Perhaps the answer to the connection problem lies in the interchangeable small grid OSS model discussed in yesterday’s article on planned OSS obsolescence.
But it probably also incorporates what ONAP calls, “real-time, policy-driven orchestration and automation,” to replace pre-defined processes. I wonder instead whether state-based transitions, being guided by intent/policy rules and feedback loops (ie learning systems) might hold the key. An evolving and learning solution that shares similarities with the electrical pathways in our brain, which strengthen the more they’re used and diminish if no longer used.

The future of work and its impact on OSS

Many years ago, I worked on a seriously big OSS transformation for one of the region’s biggest telcos. Everything was big on the project, the investment, the resources, the documentation. Everything except the outcomes. There was so much inefficiency that I often spoke about making one day of progress for every ten on site. Meetings, bureaucracy, impossible approval cycles, customer re-organisations, over-analysis, etc all added up to stagnation.

This contrasted so much with some of the amazing small teams I’ve worked alongside. Teams that worked cohesively, cleverly and just got stuff done with almost no resources. It’s one of the reasons I feel that the future of work, even for the very large organisations, will be via small teams. Outsourced to small, efficient teams / organisations. The gig economy, and the proliferation of tools that support it, make it an obvious approach to take, especially for very large organisations to leverage. Proof of work technologies, such as those building upon the discovery of blockchain, will provide further impetus to use smaller teams of experts.

Experts like a friend and colleague of mine who once built an alarm management tool in a weekend, by himself. It also happened to be more sophisticated than his employer’s existing tool that had taken years of combined developer effort by a larger team.

Maybe I’ll be proven wrong, but I see the transition to this model of work as being inevitable. The question I have is how to make our OSS more accommodating of this work model. Behemoth OSS stacks won’t. Highly modular OSS made up of many smaller components probably will, as long as they don’t succumb to the OSS chessboard analogy. The pulleys and strings will make it impossible for small, interchangable teams to decipher and manage.

A small-grid OSS model is the one I’d be backing in.

OSS – like a duck on a pond

Let’s start with a basic question. “What does an OSS need to do?”

The basic answer is, “make operations easier.”

The real answer(s) is so much more nuanced than that of course. The term easier can also encapsulate other words such as faster, more accurate, more repeatable, cheaper, etc.

Designing, building, operating and maintaining a sizable network is extremely challenging, despite network operators around the world, and the vendors that supply to them, employing some of the best and brightest. So we design OSS and related tools / processes to make operations easier.

Yet I sometimes wonder whether we achieve that aim – to make operations easier. Seems to me that we tend to focus more on just replicating functions at a higher layer in the management stack. That is, moving the function to the OSS rather than EMS/NMS, without really making it much easier operationally.

Let’s start at the user interface (UI). How often are they intuitive enough for an experienced network operator to start doing tasks with negligible OSS expert guidance?
Let’s look at deployments. How often are the projects low on effort, risk, cost and complexity?
Let’s look at flexibility (ie in-flight modifications or transformations). How often do we actually deliver flexibility to our customers through our OSS. To ask the same as above, how often are our changes low on effort, risk, cost and complexity?

As a small step towards providing an answer, I wonder whether it’s a case of making the hard things look easy and the easy things look hard.

We want to make the really hard operational things much easier to do within an OSS because that’s the primary purpose of an OSS. That’s the example of a duck on a pond. The OSS is gliding along effortlessly across the top of the water, but under the water it is paddling furiously.

Conversely, we want to make the really easy* operational things look hard to do within an OSS so that we’re not constantly being asked to build functionality / complexity into our OSS that doesn’t warrant being there. It diffuses the intent of the OSS. Just because we can, doesn’t mean we should.

Build an OSS and they will come… or sometimes not

Build it and they will come.

This is not always true for OSS. Let me recount a few examples.

The project team is disconnected from the users – The team that’s building the OSS in parallel to existing operations doesn’t (or isn’t able to) engage with the end users of the OSS. Once it comes time for cut-over, the end users want to stick with what they know and don’t use the shiny new OSS. From painful experience I can attest that stakeholder management is under-utilised on large OSS projects.

Turf wars – Different groups within a customer are unable to gain consensus on the solution. For example, the operational design team gains the budget to build an OSS but the network assurance team doesn’t endorse this decision. The assurance team then decides not to endorse or support the OSS that is designed and built by the design team. I’ve seen an OSS worth tens of millions of dollars turned off less than 2 years after handover because of turf wars. Stakeholder management again, although this could be easier said than done in this situation.

It sounded like a good idea at the time – The very clever OSS solution team keeps coming up with great enhancements that don’t get used, for whatever reason (eg non fit-for-purpose, lack of awareness of its existence by users, lack of training, etc). I’ve seen a customer that introduced over 500 customisations to an off-the-shelf solution, yet hundreds of those customisations hadn’t been touched by users within a full year prior to doing a utilisation analysis. That’s right, not even used once in the preceding 12 months. Some made sense because they were once-off tools (eg custom migration activities), but many didn’t.

The new OSS is a scary beast – The new solution might be perfect for what the customer has requested in terms of functionality. But if the solution differs greatly from what the operators are used to, it can be too intimidating to be used. A two-week classroom-based training course at the end of an OSS build doesn’t provide sufficient learning to take up all the nuances of the new system like the operators have developed with the old solution. Each significant new OSS needs an apprenticeship, not just a short-course.

It’s obsolete before it’s finishedOSS work in an environment of rapid change – networks, IT infrastructure, organisation models, processes, product offerings, regulatory shifts, disruptive innovation, etc, etc. The longer an OSS takes to implement, the greater the likelihood of obsolescence. All the more reason for designing for incremental delivery of business value rather than big-bang delivery.

What other examples have you experienced where an OSS has been built, but the users haven’t come?

Falsely rewarding based on OSS existence rather than excellence

There’s a common belief that most jobs see people rewarded for presence rather than performance. That is, they’re encouraged to be on site from 9am to 5pm rather than being given free reign over their work schedules as long as key outcomes are met / exceeded.

In OSS vendor / product selection there’s a similar concept. Contracts are often awarded based on existence rather than excellence. When evaluating a product, if it’s able to do a majority of the functions in the long list of requirements then the box is ticked.

However, this doesn’t take into account that there are usually only a very small number of functions that any given customer’s OSS needs to perform at a very high level of efficiency. All the others are effectively just nice to have. That’s the 80/20 rule at work.

When guiding a customer through their vendor selections, I always take them through an exercise to identify the use-cases / functions that really matter. Then we ensure that the demos or proofs of concept focus closely on how excellent the OSS is at those most important factors.

OSS automations – just because we can, doesn’t mean we should

Automation is about using machines / algorithms to respond faster than humans can, or more efficiently than humans can, or more accurately than humans can… but only if the outcomes justify the costs. When it comes to automations, it’s a case of, “just because we can, doesn’t mean we should.”

The more complex the decision tree you’re trying to automate, the higher the costs and therefore the harder it becomes to cost-justify. So the first step in any automation is taking a lateral thinking approach to simplifying the decision tree.

This recent post highlighted a graph from Nokia’s Bell Labs and the financial dependency that network slicing has on operational automation:
Nokia Network Slicing

Let’s use the Toyota Five Whys technique to work our way through the implications of this:

Statement 0: As CSPs, we need to drastically reduce complexity in the processes / decision-trees across our whole organisation.

Why 1? So that we can apply significant levels of automation

Why 2? So that we can apply technologies / techniques such as network slicing or virtualisation that are cost-justifiable

Why 3? So that we can offer differentiated, premium services

Why 4? So that our offerings don’t become commodities

Why 5? So that we retain corporate profitability to return to shareholders and/or invest in further interesting projects

I love that we’re looking to all number of automation technologies / techniques to apply to our OSS. However, we’re bypassing the all-important statement 0. We’re starting at Why 1 and partially missing the cost-justifiable part of Why 2. If our automation projects don’t prove cost-justifiable, then we never get the chance to reach whys 3, 4 and 5.

OSS implementation, but without the dependencies

One of the challenges with getting a new OSS or OSS transformation project completed can be the large number of dependencies that can cause momentum gridlock. If you’re looking to deliver business value in one big-bang, which is a really common approach to delivering OSS projects, then you end up juggling many different activities and hoping they all align at the right times.

I’ve noticed that the vendors tend to design their delivery schedules around big-bang / waterfall approaches like below.
Big-bang OSS delivery

Many vendors will even assure you that this is their standard practice and are hesitant to consider changes to their “best practice” delivery scheduling. Having been involved in many of these types of deliveries in the past, on both vendor and customer side, I can assure you that they rarely work well.

Generally speaking, the gridlocks occur on the customer-side, but the result is detrimental to customer and vendor alike. Hold-ups mean inefficient allocation of resources as well as the resultant cost / time over-runs.

The alternative is to apply a bit more lateral thinking to how you break down the work into smaller chunks. The lateral thinking work breakdown aims are two-fold:

  1. How to break up the work so that it best avoids dependencies; whilst also
  2. Delivering some sort of value to the customer

There are many dependencies on a typical OSS project – hardware, procurement, IT infrastructure, network connectivity, security, approvals, integrations, licensing, resource availability, data quality and many more. However, each different customer, their org chart and project has its own unique mix of dependencies, so I don’t subscribe to the “best practice” argument to project delivery.

The diagram below shows an example of an alternate breakdown. The business value chunks that are delivered might be tiny in some cases, but at least momentum can be demonstrated. Rather than having a mass of entwined dependencies, you can isolate and minimise dependencies for that sliver of business value. When the dependency/ies has cleared, you can jump straight onto the next activity from an existing build-state rather than having to align all the activities to land in perfect precision.
Incremental OSS work breakdown

OSS project stalled? Cancel it

When a project appears to be in limbo, in a permanent holding pattern, where sunk costs meet opportunity costs, where no one can figure out what to do…

Cancel it.
Cancel it with a week’s notice.

One of two things will happen:
A. A surge of support and innovation will arrive, and it won’t be stuck any more.
B. You’ll follow through and cancel it, and you won’t be stuck any more.

It costs focus and momentum to carry around the stalled. Let it go.”
Seth Godin on his blog here.

OSS projects have a tendency to get so big and complex and with so many dependencies that they can stagnate. When projects stagnate, we have a tendency of treating them with contempt or cynicism don’t we? We treat them this way even when we’re involved, so you know that outsiders are treating them with even more contempt and cynicism.

So Seth’s concept is an interesting one. I haven’t tried his technique before.

Have you? Did it achieve your desired outcomes?
Did it rally the troops? Did it clear the way for assignment of resources onto better projects, Darwinian-style? Or did it just throw away the last vestiges of momentum and all sunk costs?

OSS designed as a bundle, or bundled after?

Over the years I’m sure you’ve seen many different OSS demonstrations. You’ve probably also seen presentations by vendors / integrators that have shown multiple different products from their suite.

How integrated have they appeared to you?

  1. Have they seemed tightly integrated, as if carved from a single piece of stone?
  2. Or have they seemed loosely integrated, a series of obviously different stones joined together with some mortar?
  3. Or perhaps even barely associated, a series of completely different objects (possibly through product acquisition) branded under a common marketing name?

There are different pros and cons with each approach. Tight integration possibly suits a greenfields OSS. Looser integration perhaps better suits carve-off for best-of-breed customer architecture models.

I don’t know about you, but I always prefer to be given the impression that an attempt has been made to ensure consistency in the bundling. Consistency of user-interface, workflow, data modelling/presentation, reports, etc. With modern presentation layers, database technologies and the availability of UX / CX expertise, this should be less of a hurdle than it has been in the past.

If ONAP is the answer, what are the questions?

ONAP provides a comprehensive platform for real-time, policy-driven orchestration and automation of physical and virtual network functions that will enable software, network, IT and cloud providers and developers to rapidly automate new services and support complete lifecycle management.
By unifying member resources, ONAP is accelerating the development of a vibrant ecosystem around a globally shared architecture and implementation for network automation–with an open standards focus–faster than any one product could on its own
.”
Part of the ONAP charter from onap.org.

The ONAP project is gaining attention in service provider circles. The Steering Committee of the ONAP project hints at the types of organisations investing in the project. The statement above summarises the mission of this important project. You can bet that the mission has been carefully crafted. As such, one can assume that it represents what these important stakeholders jointly agree to be the future needs of their OSS.

I find it interesting that there are quite a few technical terms (eg policy-driven orchestration) in the mission statement, terms that tend to pre-empt the solution. However, I don’t feel that pre-emptive technical solutions are the real mission, so I’m going to try to reverse-engineer the statement into business needs. Hopefully the business needs (the “why? why? why?” column below) articulates a set of questions / needs that all OSS can work to, as opposed to replicating the technical approach that underpins ONAP.

Phrase Interpretation Why? Why? Why?
real-time The ability to make instantaneous decisions Why1: To adapt to changing conditions
Why2: To take advantage of fleeting opportunities or resolve threats
Why 3: To optimise key business metrics such as financials
Why 4: As CSPs are under increasing pressure from shareholders to deliver on key metrics
policy-driven orchestration To use policies to increase the repeatability of key operational processes Why 1: Repeatability provides the opportunity to improve efficiency, quality and performance
Why 2: Allows an operator to service more customers at less expense
Why 3: Improves corporate profitability and customer perceptions
Why 4: As CSPs are under increasing pressure from shareholders to deliver on key metrics
policy-driven automation To use policies to increase the amount of automation that can be applied to key operational processes Why 1: Automated processes provide the opportunity to improve efficiency, quality and performance
Why 2: Allows an operator to service more customers at less expense
Why 3: Improves corporate profitability and customer perceptions
physical and virtual network functions Our networks will continue to consist of physical devices, but we will increasingly introduce virtualised functionality Why 1: Physical devices will continue to exist into the foreseeable future but virtualisation represents an exciting approach into the future
Why 2: Virtual entities are easier to activate and manage (assuming sufficient capacity exists)
Why 3: Physical equipment supply, build, deploy and test cycles are much longer and labour intensive
Why 4: Virtual assets are more flexible, faster and cheaper to commission
Why 5: Customer services can be turned up faster and cheaper
software, network, IT and cloud providers and developers With this increase in virtualisation, we find an increasingly large and diverse array of suppliers contributing to our value-chain. These suppliers contribute via software, network equipment, IT functions and cloud resources Why 1: CSPs can access innovation and efficiency occurring outside their own organisation
Why 2: CSPs can leverage the opportunities those innovations provide
Why 3: CSPs can deliver more attractive offers to customers
Why 4: Key metrics such as profitability and customer satisfaction are enhanced
rapidly automate new services We want the flexibility to introduce new products and services far faster than we do today Why 1: CSPs can deliver more attractive offers to customers faster than competitors
Why 2: Key metrics such as market share, profitability and customer satisfaction are enhanced as well as improved cashflow
support complete lifecycle management The components that make up our value-chain are changing and evolving so quickly that we need to cope with these changes without impacting customers across any of their interactions with their service Why 1: Customer satisfaction is a key metric and a customer’s experience spans the entire lifecyle of their service.
Why 2: CSPs don’t want customers to churn to competitors
Why 3: Key metrics such as market share, profitability and customer satisfaction are enhanced
unifying member resources To reduce the amount of duplicated and under-synchronised development currently being done by the member bodies of ONAP Why 1: Collaboration and sharing reduces the effort each member body must dedicate to their OSS
Why 2: A reduced resource pool is required
Why 3: Costs can be reduced whilst still achieving a required level of outcome from OSS
vibrant ecosystem To increase the level of supplier interchangability Why 1: To reduce dependence on any supplier/s
Why 2: To improve competition between suppliers
Why 3: Lower prices, greater choice and greater innovation tend to flourish in competitive environments
Why 4: CSPs, as customers of the suppliers, benefit
globally shared architecture To make networks, services and support systems easier to interconnect across the global communications network Why 1: Collaboration on common standards reduces the integration effort between each member at points of interconnect
Why 2: A reduced resource pool is required
Why 3: Costs can be reduced whilst still achieving interconnection benefits

As indicated in earlier posts, ONAP is an exciting initiative for the CSP industry for a number of reasons. My fear for ONAP is that it becomes such a behemoth of technical complexity that it becomes too unwieldy for use by any of the member bodies. I use the analogy of ATM versus Ethernet here, where ONAP is equivalent to ATM in power and complexity. The question is whether there’s an Ethernet answer to the whys that ONAP is trying to solve.

I’d love to hear your thoughts.

(BTW. I’m not saying that the technologies the ONAP team is investigating are the wrong ones. Far from it. I just find it interesting that the mission is starting with a technical direction in mind. I see parallels with the OSS radar analogy.)

Where are the reliability hotspots in your OSS?

As you already know, there are two categories of downtime – unplanned (eg failures) and planned (eg upgrades / maintenance).

Planned downtime sounds a lot nicer (for operators) but the reality is that you could call both types “incidents” – they both impact (or potentially impact) the customer. We sometimes underestimate that fact.

Today’s question is whether you’re able to identify where the hotspots are in your OSS suite when you combine both types of downtime. Can you tell which outages are service-impacting?

In a round-about way, I’m asking whether you already have a dashboard that monitors uptime of all the components (eg applications, probes, middleware, infra, etc) that make up your complete OSS / BSS estate? If you do, does it tell you what you anecdotally know already, or are there sometimes surprises?

Does the data give you the evidence you need to negotiate with the implementers of problematic components (eg patch cadence, the need for reliability fixes, streamlining the patch process, reduction in customisations, etc)? Does it give you reason to make architectural changes (eg webscaling)?

Persona mapping for OSS PoCs

When selecting new applications for an OSS or to augment an existing OSS, it always makes sense to me to run a Proof of Concept. But what do we want to demonstrate in that PoC? For me, we want to run demonstrations of the factors (eg features, use-cases, processes, etc) that justify the investment.

A simple exercise you can use is to identify the personas / roles that interact with the OSS. This could include personas such as NOC operator, strategic planner, network engineer, order entry, field ops, data / analytics, application administrator, etc. The actual personas will differ within each organisation of course.

For each of those personas, we can identify and interview an individual that represents that persona.

Interview questions include:

  1. What are the key responsibilities of your role
  2. What is the most important goal / KPI for your role
  3. How does this OSS (or proposed OSS) support you meeting this goal
  4. Describe the single most important process / function that you perform using the OSS
  5. Why is it so important
  6. How often do you perform this process / function
  7. Please provide a short list of other important processes / functions you perform with this OSS

We can then build this into a matrix and seek to prioritise into a set of use-cases. Based on time and cost constraints, we can then build the top-n of those use-cases into implementation scenarios for the PoC.

OSS operationalisation at scale

We had a highly flexible network design team at a previous company. Not because we wanted to necessarily, but because we were forced to by the client’s allocation of work.

Our team was largely based on casual workers because there was little to predict whether we needed 2 designers or 50 in any given week. The workload being assigned by the client was incredibly lumpy.

But we were lucky. We only had design work. The lumpiness in design effort flowed down through the work stack into construction, test and deployment teams. The constructors had millions of dollars of equipment that they needed to mobilise and demobilise as the work ebbed and flowed. Unfortunately for the constructors, they’d prepared their rate cards on the assumption of a fairly consistent level of work coming through (it was a very big project).

This lumpiness didn’t work out for anyone in the delivery pipeline, the client included. It was actually quite instrumental in a few of the constructors going into liquidation. The client struggled to meet roll-out targets.

The allocation of work was being made via the client’s B/OSS stack. The B/OSS teams were blissfully unaware of the downstream impact of their sporadic allocation of designs. Towards the end of the project, they were starting to get more consistent and delivery teams started to get into more of a rhythm… just as the network was coming to the end of build.

As OSS builders, we sometimes get so wrapped up in delivering functionality that we can forget that one of the key requirements of an OSS is to operationalise at scale. In addition to UI / CX design, this might be something as simple as smoothing the effort allocation for work under our OSS‘s management.

Help needed: IoT / OSS cross-over use cases

Hi PAOSS community.

I’d like to call in a favour today if I may. I’m on the hunt for any existing use-cases and / or project sites that have integrated a significant sensor network into their OSS and existing operational processes.

That includes a strategy for handling IoT-scale integration of data collection, event / alarm processing, device management, data contextualization, data analytics, end-to-end security and applications management / enablement within existing OSS tools.

I’m looking for examples where an OSS had previously managed thousands of (network) devices and is now managing hundreds of thousands of (IoT) devices. Not necessarily IoT devices of customers as services but within an operator’s own network.

Obviously that’s an unprecedented change in scale in traditional OSS terms, but will be commonplace if our OSS are to play a part in the management of large sensor networks in the future.

There’s an element of mutual exclusivity between what an IoT management platform and OSS needs to do, but there are also some similarities. I’d love to speak with anyone who has actually bridged the gap.

What OSS environments do you need?

When we’re planning a new OSS, we tend to be focused on the production (PROD) environment. After all, that’s where it’s primary purpose is served, to operationalise a network asset. That is where the majority of an OSS‘s value gets created.

But we also need some (roughly) equivalent environments for separate purposes. We’ll describe some of those environments below.

By default, vendors will tend to only offer licensing for a small number of database instances – usually just PROD and a development / test environment (DEV/TEST). You may not envisage that you will need more than this, but you might want to negotiate multiple / unlimited instances just in case. If nothing else, it’s worth bringing to the negotiation table even if it gets shot down because budgets are tight and / or vendor pricing is inflexible relating to extra environments.

Examples where multiple instances may be required include:

  1. Production (PROD) – as indicated above, that’s where the live network gets managed. User access and controls need to be tight here to prevent catastrophic events from happening to the OSS and/or network
  2. Disaster Recovery (DR) – depending on your high-availability (HA) model (eg cold standby, primary / redundant, active / active), you may require a DR or backup environment
  3. Sandpit (DEV / TEST) – these environments are essential for OSS operators to be able to prototype and learn freely without the risk of causing damage to production environments. There may need to be multiple versions of this environment depending on how reflective of PROD they need to be and how viable it is to take refresh / updates from PROD (aka PROD cuts). Sometimes also known as non-PROD (NP)
  4. Regression testing (REG TEST) – regression testing requires a baseline data set to continually test and compare against, flagging any variations / problems that have arisen from any change within the OSS or networks (eg new releases). This implies a need for data and applications to be shielded from the constant change occurring on other types of environments (eg DEV / TEST). In situations where testing transforms data (eg activation processes), REG TEST needs to have the ability to roll-back to the previous baseline state
  5. Training (TRAIN) – your training environments may need to be established with a repeatable set of training scenarios that also need to be re-set after each training session. This should also be separated from the constant change occurring on dev/test environments. However, due to a shortage of environments, and the relative rarity of training needed at some customers, TRAIN often ends up as another DEV or TEST environment
  6. Production Support (PROD-SUP) – this type of environment is used to prototype patches, releases or defect fixes (for defects on the PROD environment) prior to release into PROD. PROD-SUP might also be used for stress and volume testing, or SVT may require its own environment
  7. Data Migration (DATA MIG) – At times, data creation and loading needs to be prototype in a non-PROD environment. Sometimes this can be done in PROD-SUP or even a DEV / TEST environment. On other occasions it needs its own dedicated environment so as to not interrupt BAU (business as usual) activities on those other environments
  8. System Integration Testing (SIT)OSS integrate with many other systems and often require dedicated integration testing environments

Am I forgetting any? What other environments do you find to be essential on your OSS?

The OSS self-driving vehicle

I was lucky enough to get some time of a friend recently, a friend who’s running a machine-learning network assurance proof-of-concept (PoC).

He’s been really impressed with the results coming out of the PoC. However, one of the really interesting factors he’s been finding is how frequently BAU (business as usual) changes in the OSS data (eg changes in naming conventions, topologies, etc) would impact results. Little changes made by upstream systems effectively invalidated baselines identified by the machine-learning engines to key in on. Those little changes meant the engine had to re-baseline / re-learn to build back up to previous insight levels. Or to avoid invalidating the baseline, it would require re-normalising all of data prior to the identification of BAU changes.

That got me wondering whether DevOps (or any other high-change environment) might actually hinder our attempts to get machine-led assurance optimisation. But more to the point, does constant change (at all levels of a telco business) hold us back from reaching our aim of closed-loop / zero-touch assurance?

Just like the proverbial self-driving car, will we always need someone at the wheel of our OSS just in case a situation arises that the machines hasn’t seen before and/or can’t handle? How far into the future will it be before we have enough trust to take our hands off the OSS wheel and let the machines drive closed-loop processes without observation by us?

An alternate way of slicing OSS (part 2)

Last week we talked about an alternate way of slicing OSS projects. Today, we’ll look a little deeper and include some diagrams.

The traditional (aka waterfall) approach to delivering an OSS project sees one big-bang delivery of business value at the end of the implementation.
OSS project delivery via waterfall

The yellow arrows indicate the sequential nature of this style of delivery. The implications include:

  1. If the project runs out of funds before the project finishes, no (negligible) value is delivered
  2. If there’s no modularity of delivery then the project team must stay the course of the original project plan. There’s no room for prioritising or dropping or including delivery modules. Project plans are rarely perfect at first after all
  3. Any changes in project plan tend to have knock-on effects into the rest of the delivery
  4. There is only one true delivery of value, but milestones demonstrate momentum for the project… a key for change management and team morale
  5. Large deliverables represent the proverbial overload one segment of the project delivery team then under-utilises the rest in each stage.  This isn’t great for project flow or team utilisation

The alternate approach seeks to deliver in multiple phases by business value, not artefacts, as shown in the sample model below:
OSS project delivery via AgilePhased enhancements following a base platform build (eg Sandpit and/or Single-site above) could include the following, where each provides a tangible outcome / benefit for the business, thus maintaining perception of momentum (assurance use-cases cited):

  • Additional event collection (ie additional collectors / probes / mediation-devices can be added or configured)
  • Additional filters / sorting of events
  • Event prioritisation mapping / presentation
  • Event correlation
  • Fault suppression
  • Fault escalation
  • Alarm augmentation
  • Alarm thresholding
  • Root-cause analysis (intra, then inter-domain)
  • Other configurations such as latching, auto-acknowledgement, visualisation parameters, etc
  • Heart-beat function (ie devices are unreachable for a user-defined period)
  • Knowledge base (ie developing a database of activities to respond to certain events)
  • Interfacing with other systems (eg trouble-ticket, work-force management, inventory, etc)
  • Setup of roles/groups
  • Setup of skills-based routing
  • Setup of reporting
  • Setup of notifications (eg email, SMS, etc)
  • Naming convention refinements
  • etc, etc

The latter is a more Agile-style breakdown of work, but doesn’t need to be delivered using Agile methodology.

Of course there are pros and cons of each approach. I’d love to hear your thoughts and experiences with different OSS delivery approaches.

Optimisation Support Systems

We’ve heard of OSS being an acronym for operational support systems, operations support systems, even open source software. I have a new one for you today – Optimisation Support Systems – that exists for no purpose other than to drive a mindset shift.

I think we have to transition from “expectations” in a hype sense to “expectations” in a goal sense. NFV is like any technology; it depends on a business case for what it proposes to do. There’s a lot wrong with living up to hype (like, it’s impossible), but living up to the goals set for a technology is never unrealistic. Much of the hype surrounding NFV was never linked to any real business case, any specific goal of the NFV ISG.”
Tom Nolle
in his blog here.

This is a really profound observation (and entire blog) from Tom. Our technology, OSS included, tends to be surrounded by “hyped” expectations – partly from our own optimistic desires, partly from vendor sales pitches. It’s far easier to build our expectations from hype than to actually understand and specify the goals that really matter. Goals that are end-to-end in manner and preferably quantifiable.

When embarking on a technology-led transformation, our aim is to “make things better,” obviously. A list of hundreds of functional requirements might help. However, having an up-front, clear understanding of the small number of use cases you’re optimising for tends to define much clearer goal-driven expectations.

Security and privacy as an OSS afterthought?

I often talk about OSS being an afterthought for network teams. I find that they’ll often design the network before thinking about how they’ll operationalise it with an OSS solution. That’s both in terms of network products (eg developing a new device and only thinking about building the EMS later), or building networks themselves.

It can be a bit frustrating because we feel we can give better solutions if we’re in the discussion from the outset. As OSS people, I’m sure you’ll back me up on this one. But we can’t go getting all high and mighty just yet. We might just be doing the same thing… but to security, privacy and analytics teams.

In terms of security, we’ll always consider security-based requirements (usually around application security, access management, etc) in our vendor / product selections. We’ll also include Data Control Network (DCN) designs and security appliance (eg firewalls, IPS, etc) effort in our implementation plans. Maybe we’ll even prescribe security zone plans for our OSS. But security is more than that (check out this post for example). We often overlook the end-to-end aspects such central authentication, API hardening, server / device patching, data sovereignty, etc and it then gets picked up by the relevant experts well into the project implementation.

Another one is privacy. Regulations like GDPR and the Facebook trials show us the growing importance of data privacy. I have to admit that historically, I’ve been guilty on this one, figuring that the more data sets I could stitch together, the greater the potential for unlocking amazing insights. Just one problem with that model – the more data sets that are stitched together, the more likely that privacy issues arise.

We increasingly have to figure out ways to weave security, privacy and analytics into our OSS planning up-front and not just think of them as overlays that can be developed after all of our key decisions have been made.

An alternate way of slicing OSS projects

One of the biggest challenges of big bang OSS project implementations is that all of the business value (ie the OSS and its data, workflows, integrations, etc) gets delivered at once, normally at the end of a lengthy exercise.

Ok, ok, so the delivery of value is not a challenge, it’s the implications of a big delivery of value that’s the challenge – implications that include:

  1. If the project runs out of funds before the project finishes, no value is delivered
  2. If there’s no modularity of delivery then the project team must stay the course of the original project plan. There’s no room for prioritising or dropping or including delivery modules. Project plans are rarely perfect at first after all
  3. Any changes in project plan tend to have knock-on effects into the rest of the delivery due to the sequential nature of typical project plans
  4. Any delivery of value represents a milestone, which in turn demonstrates momentum for the project… a key change management and team morale strategy
  5. Large deliverables represent the proverbial “pig in the python” – only one segment of the python (ie segment of the project delivery team) is engaged (hyper-engaged) whilst the other segments remain under-utilised.  This isn’t great for project flow or utilisation

When tasked with designing a project schedule, I’ve noticed that many vendors tend to follow the typical waterfall delivery and corresponding payment milestones (eg. design, then build, then test, then deploy, then hand over). The downside of this approach is that the business value (for the customer) is delivered at the end of the handover (ie big bang). There’s no business value in delivering design artefacts for example – the customer can’t use them to perform operational tasks.

The model I prefer sees incremental business value being delivered such as:

  • Proof of Concept (PoC) build
  • Sandpit build
  • Out of the box (OOTB) production build (ie. no customisations)
  • End-to-end use case #1 delivery (ie. design, build*, test, deploy, handover)
  • E2E use case #2 delivery
  • E2E use case #n delivery

where build* includes incremental configuration, customisation, integration, data migration, etc.