Zoe Food Health App Experience

About 2 weeks ago I received my Zoe test kit.

The yellow box that was delivered by the Royal Mail was overkill in my view. An expense that after a number of days, felt like Zeo had concentrated on the wrong parts of the overall Customer User Experience.

First, let me summaries the positives from my experience:

• On-line chat support was helpful
• Libre app and sensor for collecting blood glucose readings is easy to use, and the biggest value add in experimenting with food
• Blood collection is relatively pain free

The unfortunate negatives:

• If anything goes wrong on the order of your tasks day to day in the Zoe App, Zoe support can’t do a lot to correct the order, impacting the overall User Experience within the app, and causing frustration
• Using two apps, Zoe and Libre was just wrong from a User Experience
• When filling in questionnaires in the app, there appears to be no way to pause, and resume. Just discard and start again – challenging
• There is some duplication on data entry across the initial week of questionnaires
• The email “blood sugar challenge plan” is sent from the domain typeform.com
• There is very limited feedback and insight in the first few weeks, which comes back again to overall poor User Experience

In summary, it was an interesting experience. For anyone who has already spent time reading about nutrition, you maybe best to start with Libre app and then consider if Zoe is worth the $

UDM Pro connected to Plusnet and Hey!Broadband

Having just spent what feels like years waiting for fibre to be available in my area of the woods, Three weeks ago, I’ve finally had the fibre installed into my house but took a few weeks to resolve the connectivity issues 😦 Here’s a summary to help anyone else following in my footsteps:

I’ve been running with Plusnet broadband for a long time. The broadband comes in on a copper phone line, into an OpenReach bridge, and from the bridge to a Plusnet 4 port router. Previously I had the router connected into Port9 (default) of the UDM Pro (WAN). This works fine, until you decide you want a static IP address, at which point you need to connect the OpenReach bridge directly into WAN which is the default Port9, and go via the UDM Pro dashboard, settings/internet, select Default (WAN) and in the advance section, select manual, and for IPv4 Connection select PPPoE, and sent the Plusnet username and password. This should allow the UDM Pro to PPPoE to Plusnet.

I made the decision to keep Plusnet for a period of time whilst I get Hey!Broadband Fibre install, and get a bearing on the quality of service. Plusnet has delivered a good quality of service over the last n years, but doesn’t offer fibre in my area.

Hey!Broadband Fibre cones into a bridge model, which in non bridge install, would connect to a wifi router. I opted for bridge mode, with a static IP, and therefore didn’t need the wifi router. UDM Pro Port10 (WAN2) is the port that requires the ethernet cable from Fibre bridge modem. Once this is connected, UDM Pro dashboard, settings/internet, select Backup (WAN2) advance settings needs to have PPPoE enabled, with the appropriate username and password. If you want to view the logs whilst the PPPoE is enabled, SSH into the UDM Pro from any machine on your LAN, and tail /var/log/message. You may also want to ifconfig and route -n to check the interface is setup as expected on a successful connection. Errors I experience during the initial setup of the service ranged from failing to find remote host, Unsupported address family, Misformatted DNS reply and Unable to complete PPPoE Discovery

Hopefully by now your UDM Pro dashboard (top left) should display WAN and WAN2 with IP addresses above the Gateway IP. You should be able to run Speed Tests for both internet connections (WAN/WAN2) from the UDM Pro dashboard (bottom right), by selecting WAN or WAN2 and then selecting Speed Test.

I ran into two issue during the overall install of fibre. Hey!Broadband hasn’t setup the bridge modem correctly. I diagnosed this by using a laptop, and PPPoE’d to the bridge modem, and reporting the issues I saw. Post a successful PPPoE connecting, I saw DNS resolution issues on the bridge modem or further back into Hey!Broadband’s network. Reporting this to TechSupport resolve the issue at 6.40am the following morning – effectively they assigned me a new static IP

Between a successful PPPoE connection, and resolved DNS issues, I wanted the OpenReach PPPoE to work from Port10 (WAN2), since it was going to be the failover. The reason for this turned out to be that Port10 can’t downgrade to 100Mb, which is the max connection the OpenReach bridge can achieve. Effectively I now had Port9 (WAN) Plusnet, and Port10(WAN2) Hey!Broadband Fibre, which is the wrong way around from a failover, and default UDM Pro setup. To resolve this I went into the UDM Pro Port Management, and swapped WAN to Port10 and WAN2 to Port9. Solved

With the above failover configuration in place, PPPoE connected, but the DNS issues unresolved, the UDM Pro failed over to WAN2, and the slower Plusnet broadband. Sometimes between 1am and 6:40am, Hey!Broadband resolve the DNS/PPPoE issues, and UDM Pro failed back to WAN.

2022 Rules of Engagement

Following on from 2006/2013, here’s the updated set:

1. Be yourself
2. Lead from the front
3. Code == source of truth
4. Observability from day 1
5. Working backwards with tenets and mechanisms
6. Customer centric UX Design (Engineers are user too)
7. Team efficiency is important
8. Measure and adapt
9. Behaviour changes culture
10. Keep the message simple, and repeat
11. Simple over complexity
12. Mandates often have consequences. Autonomy needs an economy

Team Productivity

As I read Project to Product by Mik Kersten, I’m reminded of the need for teams to remain durable and responsible as part of their journey:

Giving work to an existing high-performing team will get you results much faster than having to form a team first, then dissolve that team at the end.

The alarm in the above quote is “dissolve”. Team are so often dissolved before they perform. In my career, organisations seem to rarely discuss the impact of mobility/terminations on the team, possibly causing a cycle of forming/storing which can make it to performing. Scaling of people to aid a project, is often considered in terms of adding budget to increase headcount, forgetting about the need to consider product over project. The counter to this is that headcount doesn’t guarantee an increase in velocity. Brining us nearly to flow and value stream.

As DORA and SPACE offer a measure of engineering productivity, we need to pause and consider the customer experience of productivity rather than just the engineering, thereby offer insight into what business value you’re delivering, and how quickly. Flow of value.

This leads us to the need for organisation change to allow team to become effective, which in itself is a considerable challenge, since we are probably all aware of the ivory towers that exist. McKinsey offer some insight into the journey

how do I accomplish my 10 year plan in 6 months

SRE Workbook Notes

A few items I found insightful, outside of an overall great book:

• Automate This Year’s Job Away
• VALET Dashboard – like the ideas.
• Using VALET in Testing
• Testing Alerting Logic
• Identify and Measure Toil
• Zombie Code Slayer badge
• Main Roles in Incident Response
• Gamification
• Jsonnet – Configuration language
• Release Engineering Principles
• Production Readiness Review
• SRE Practices Without SREs – Start TODAY!

Engineering Company = Engineering Culture

Great article on Wall St vs New Order/Startup companies.  One particular statement stands out:

Engineering Company = Engineering Culture
In Wall Street, Finance is the business, and technology is merely a mechanism that exists to facilitate that business. Technology is seen as a cost center, and when it is necessary to cut costs

SRE Post Mortem Gamification

I really like gamification.  In some firms its in almost part of the culture.  Of particular interest around post mortem’s is Google’s gamification leaderboard.

Some individuals are incentivized by a sense of accomplishment and progress toward a larger goal, such as fixing system weaknesses and increasing reliability. For these individuals, a scoreboard or burndown of postmortem action items can be an incentive. At Google, we hold “FixIt” weeks twice a year. SREs who close the most postmortem action items receive small tokens of appreciation and (of course) bragging rights

Cyber Security – Notes

A few weeks ago, the Sunday Times included a CyberSecurity supplement, available online here.  What follows are a few interested pointers:

• Human Centered Security
• Phishing attempts is number one in terms of insider threat
• Increase in awareness and training can lead to the biggest improvement in security
• Cloud – understand where your sensitive data is, assess access and sharing privileges.
• Open source vulnerabilities create serious risks.

DevSecOps – Secure Code

Worth a listen, Max Saltonstall and Justin McCarthy are joined by Johnathan Hunt, VP of Information Security at InVision to talk about pen testing, bug bounty programs, and secure code.

• Pen Testing yearly cycle – “significant flaw in thinking”
• “static analysis scanner that sits locally on all software engineers laptops, every piece of code every line of code that they write their supposed to scan this prior to committing that to the repos. Once it’s in the repos, once we get ready to deploy and merge in a master at that point that runs again, right, the same tool runs within a CI CD pipeline, after we’re doing, QA testing, and all these other things that run also is an automated tool set, it runs again, at that point, it notifies us or notifies them right of vulnerabilities resident, now we can choose to block that we can choose to say, hey, if it’s a critical vulnerability, or a high severity vulnerability, we’re going to disable or block the push right to production”

Software Productivity

First up, and worth a read/watch is “How Is Software Developed At Amazon?”.  Of note:

• Automate anything manual
• DevSecOps

Following closely are the many references called out in QA Financial keynote from Jordan Daniel, Head of Testing and Enterprise Release Management, Bank of England:

• Boehm, B. W. (1987) Improving Software Productivity
• Grady, R. B. (1992) Practical Software Metrics for Project Management and Process Improvement, 1st Edition Prentice Hall.
• Henderson, P. (2006) Why Large IT Projects Fail.
• Jones, T. C. (2012) Software Defect Origins and Removal Methods. Draft 5.0. p. 3.
• McConnell, S. (1996) Software Quality at Top Speed.
• McConnell, S. (1996) Rapid Development: Taming Wild Software Schedules. Microsoft Press
• McConnell, S. (2004) Code Complete: A Practical Handbook of Software Construction, 2nd Edition Microsoft Press
• Medina, L., Liedke, R. (2015) Defect Prevention Using Agile Techniques.
• Patton, R. (2005) Software Testing. 2nd Edition Sams Publishing
• Suma. V., Gopalakrishnan Nair, T. R. (2008) Effective Defect Prevention Approach in Software Process for Achieving Better Quality Levels.