Tag Archives: Velocity

Benefits of receiving a UCISA bursary

Giuseppe Sollazzo

 

 

 

Giuseppe Sollazzo
Senior Systems Analyst
St George’s, University of London

 

 

 

 

Last October I was lucky enough to be selected for a UCISA bursary to attend O’Reilly Velocity in Amsterdam. Velocity is one of the most important conferences for performances in IT Systems, which is my area of work at St George’s, University of London: I lead a team of systems analysts who take care of the ongoing maintenance and development of our infrastructure. I had wanted to attend the conference for quite a while, but was always prevented from doing so by the hefty funding required, something that my institution could not readily justify.

The format of Velocity is particularly well suited to a mixture of blue-sky thinking, practical learning, networking with other professionals. Each day ran from 8:30 till 18:30. Following this schedule for three days was intense, but extremely rewarding in terms of learning.

I have written blogs for UCISA day by day throughout the conference. You can read about the specific sessions I followed on each day at the following links: day one, day two and day three. In summary, I learned about a mixture of practical techniques and heard about experiences in a variety of sectors.

As I wrote in my first blog post ahead of the conference, a focus on performance and optimisation is important for academic IT services, and specifically for my institution: with our 300 servers and 30,000 accounts to take care of, this is not just an important consideration, but our major worry on a daily basis. Access to funding is becoming increasingly competitive, as is student and researcher recruitment; it is becoming our primary goal to provide systems that are effective, secure, scalable, fast, and at the same time manageable by constrained staff numbers.

I was interested in three types of sessions:

  • practical tutorials about established techniques and tools
  • storytelling from people who have applied techniques to certain specific situations
  • sessions about new learning about new systems, to see where the industry is heading to.

Velocity has been great to help me crystallise my strategy on how to make St George’s systems evolve. In the past four months, this has translated into taking action on a number of aspects of our infrastructure. The most important are the following:

  • leading the team to build upon our logging systems, in order to extract metrics and improve the ability to respond to incidents
  • increasing our dependability on our ticketing system, by measuring response times and starting a project to make the ongoing monitoring of this part of our weekly service reviews
  • launching an investigation into researchers’ needs in terms of data storage and high performance computing; this has so far resulted in an experimental HPC cluster, which we are testing in collaboration with genomics and statistical researchers who are interested in massively parallel computations where performances are vital to the timeliness of research results for publishing.

I’m very grateful to UCISA for the opportunity it has given me. The knowledge and experience I’ve gathered at Velocity have been invaluable not just for starting new projects and reviewing our current service offer, but most importantly in beginning to understand what our strategy to maintain performances should be to still be able, in five to ten years’ time, to provide excellent industry-standard services to our community.

Interested in applying for a UCISA bursary? Then visit UCISA Bursary Scheme 2018.

Performance management and assessing capacity

Giuseppe Sollazzo

 

 

Giuseppe Sollazzo
Senior Systems Analyst
St George’s, University of London

 

 

 

 

 

Velocity day three – the final one – has been another mind-boggling combination of technical talks and masterful storytelling about performance improvement in a disparate set of systems. The general lesson of the day is: know your user, know your organization, know your workflows – only then will you be able to adequately plan your performance management and assess your capability.

This was the message from the opening keynote by Eleanor Saitta. She spoke about how to design for ‘security outcomes’, or, in other words, ‘security for humans’: there is no threat management system that works if isolated from an understanding of the human system where the threats emerge. We have some great examples of this in academia, and at St George’s one of the major challenges we face is securing systems and data in a context of academic sharing of knowledge. Being a medical school, the human aspect of security – and how this can affect performances – is something we have to face on a daily basis.

One of the best presentations, however, was by David Booker of IBM, who gave a live demo of the Watson system, an Artificial Intelligence framework which is able to understand informal (up to a point) questions and answer them in speaking. As per every live demo, this encountered some issues. Curiously, Watson wasn’t able to understand David’s pronunciation of the simple word “yes”. “She doesn’t get when I say ‘yes’ because I’m from Brooklyn,” David said, triggering laughter in the audience.

Continuous delivery
Courtney Nash of O’Reilly spoke at length about how we should be thinking when we build IT services, with a focus on the popular strategy of continuous delivery. Continuous delivery is the idea that a system should transition from development to production very often, and this idea is taking traction in both industry and academia. However, this requires trust: trusting your tools, your infrastructure, your code, and most importantly, the people who power the whole organization. Once again, then, we see the emergence of a human factor when planning for the delivery of IT services.

The importance of 2G
In another keynote with a lot of applicable ideas for academic websites, Bruce Lawson of Opera ASA has focused on the ‘next billion’ users from developing countries who are starting to use internet services. Access to digital is spreading, especially in developing areas of Asia, where four billion people live. India had 190 million internet users in 2014, and this is poised to grow to 400 million by 2018.

The best piece of information in this talk was the realisation that if you take the US, India and Nigeria, the top 10 visited websites are the same: Facebook, Gmail, Twitter, and so on. Conversely, the top 10 devices give a very different picture: iPhones dominate in the US, cheap Androids in India, and Nokia or other regional feature phones in Nigeria. This teaches us an important lesson: regardless of hardware, people worldwide want to consume the same goods and services. This should tell us to build our services in a 2G-compatible way if we want to reach the next billion users (91.7% people in the world live within reach of a 2G network). This is of great importance to academia in terms of international student recruitment.

Performance optimisation
The afternoon sessions were an intense whistle-stop tour of experiences of performance optimisation. Alex Schoof of Fugue, for example, gave an intensely technical session about secret management in large scale systems, something that definitely applies to our context: how do we distribute keys and passwords in a secure way that allows that secrets to be changed whenever required? With security issues going mainstream, like the infamous Heartbleed bug, this is something of increasing importance. Adam Onishi of London-based dxw, a darling of public sector website development, gave an interesting talk on how performance, accessibility and technological progress in web design are interlinked, something academic website managers have too often failed to consider with websites that are published and then forgotten for years.

As someone who has developed mobile applications, I really enjoyed AT&T’s Doug Sillars’ session about ‘bad implementation of good ideas’, showing that lack of attention to the system as a whole has often killed otherwise excellent apps, which are too focused on local aspects of design.

Velocity has been a great event. I was worried it would be too ‘corporate’ or sponsor-oriented, but it has been incredibly rich, with good practical ideas that I could apply to my work immediately. It has also offered some good reflection on ‘running your systems in house’: we often perceive this dualism between the Cloud and in-house services. This is a technology that can be run in-house with no need to outsource. As IT professionals we should appreciate it, and make the case for adopting technologies that improve performance and compliance in a financially sound way. This often requires abandoning outsourcing and investing on internal resources: a good capital investment that will allow continuous improvement of the infrastructure.

 

PaaS, bots, alerts and using analytics to improve web performance

Giuseppe Sollazzo

 

 

 

Giuseppe Sollazzo
Senior Systems Analyst
St George’s, University of London

 

 

Storytelling at Velocity

The second day of O’Reilly Velocity conference was definitely about storytelling: keynotes and sessions were both descriptions of performance-enhancement projects or accounts of particular experiences in the realm of systems management, and in all honesty, many of these stories resonate with our daily experience running IT Services in an academic environment. I will give a general summary, but also mention the names of the speakers I’ve found most useful.

Evolution in the Internet of Things age
An attention-catching keynote by Scott Jenson, Google’s Physical Web project lead, the first session was centred on a curious observation: most attention about web performances has traditionally been focused on the “body”, the page itself, while the most interesting and performance-challenged part is actually the address bar.

Starting from this point, Scott has illustrated how the web is evolving and what its characteristics will be especially in the Internet of Things age. He advocated for this to be an “open” project, rather than Google’s.

Another excellent point he has made is that control should be given back to the users. This was illustrated by a comparison between a QR code and an iBeacon : the former requires the user to take action; the latter is proactive to a passive user. Although we like to think of proactive applications, it only takes us to walk into a room full of them to understand being in control can be a good thing.

PaaS for Government as a Platform
Most of the conference talks have centred on monitoring and analytics as a way to manage performances. Among the most interesting talks, Anna Shipman of the UK Government Digital Service (GDS) illustrated how they are choosing a Platform-as-a-Service supplier in order to implement their “Government-as-a-Platform” vision.

I’ve argued a lot in the past that UK Academia will need, sooner or later, to go through a “GDS moment” to get back to innovation in a way it can control – as opposed to outsource in bulk – and this talk was definitely a reminder of that.

Rise of the bot
As with yesterday’s Velocity sessions, some truly mind-boggling statistics have been released today. One example is that that many servers are overwhelmed by web crawlers or “bots” – the automated software agents that index websites for search engines. In his presentation From RUM to robot crawl experience!  Klaus Enzenhofer of Dynatrace told the audience that he spoke to several companies for which two thirds of all traffic they receive is Google Bots. “We need a data centre only for Google”, they say.

Analytics for web performance
There has been quite a lot of discussion around monitoring vs. analysis. In his presentation Analytics is the new monitoring: Anomaly detection applied to web performance Bart De Vylder of CoScale argued for the adoption of data science techniques in order to build automatic analysis procedures for smart, adaptive alerting of anomalies. This requires an understanding of the domain of the anomalies in order to plan how to evolve the monitoring, considering for example seasonal variations in web access.

Using alerts
On a similar note was the most oversubscribed talk of the day, a 40 minute session by Sarah Wells of the Financial Times which saw over 200 attendees (with many trying to get a glimpse from outside the doors). Sarah told the audience about how it is very easy to be overwhelmed by alerts: in the FT’s case, they perform 1.5M checks per day generating over 400 alerts per day. She gave an account of their experience trimming down these figures. Very interestingly, the FT has adopted the cloud as a technology, but they haven’t bought it from an external supplier: they’ve built it themselves, with great attention to performance, cost, and compliance, surely a strategy that I subscribe to.

Conference creation
I also attended an interesting non-technical session by another Financial Times employee, Mark Barnes, who explained how they conceived the idea of an internal tech conference and how they effectively run it.

Hailed an internal success and attended by their international crowd, the conference idea came from an office party and reportedly has helped improve internal communications at all levels. As a conference/unconference organiser myself (OpenDataCamp, UkHealthCamp, WhereCampEU, UKGovCamp, and more), having this insight from the Financial Times will be invaluable for future events.

I’m continuing to fill in this Google doc with technical information and links from the sessions I attend, so have a look if you’re interested.

Disruptive statistics, Linux containers, extreme web performance for mobile devices

Giuseppe Sollazzo

 

 

 

Giuseppe Sollazzo
Senior Systems Analyst
St George’s, University of London

 

 

 

 

Day one at the Velocity conference, Amsterdam

What a first day! O’Reilly Velocity, the conference I’m attending thanks to a UCISA bursary, is off to a great start with a first day oriented to practical activities and hands-on workshops. The general idea of these workshops is to build and maintain large-scale IT systems enhancing their performances. Let me provide you with a quick summary of the workshops I have attended.

Statistics for Engineers
A statistics workshop at 9.30am is something that most would find soul-destroying, but this was a great introduction on how to use statistics in an engineering context – in other words, how to apply statistics to reality in order to gather information with the goal of taking action.

Statistics is, indeed, very simple maths and its difficult yet powerful bits allow practitioners to understand situations and predict their outcomes.

This workshop illustrated how to apply statistical methods to datasets generated by user applications: support requests, server logs, website visits. Why is this important? Very simply because service levels need to be planned and agreed upon very carefully. The speaker showed some examples of this. In fact, the title of this workshop should have been “Statistics for engineers and managers”: usage statistics help allocate resources (do we need more? can we reuse some?) and, in turn, financial budgets.

The workshop illustrated how to generate descriptive statistics and also how to use several mathematical tools for forecasting the evolution of service levels. We have had some experience with data collection and evaluation at St George’s University of London, and this workshop has definitely helped refine the tools and reasoning we will be applying.

Makefile VPS
This talk presented itself as a super-geeky session about Linux containers. Containers are a popular way to manage web services that does not require a full-fledged physical or virtual server. They can be easily built, deployed, and managed. However, they are rarely properly understood.

The engineer who presented this workshop showed how in his company, SoundCloud,  they build their own containers to power a “virtual lab” in order to simulate failures and train their engineers to react. His technique, based on scripts that build and launch containers at the press of the “Enter” button, is an effective solution both for quick prototyping and production deployment whenever docker or other commercial/free solutions are not a viable option (due to funding or complexity).

As much as this was quite a hard core session, it was good to see how services can be run in a way that makes their performances very easy to manage. This is definitely something that I will be sharing with my IT colleagues.

Extreme web performance for mobile devices
A lightweight (so to say!) finale to the day, discussing how mobile websites present a diverse range of performance issues and what techniques can be used to test and improve. However, the major contribution from this session was to share some truly extraordinary statistics about mobile traffic and browsers.

For example, the fact that on mobile 75% of traffic is from browser and 25% from web views (i.e. from apps) – 40% of which is from Facebook. Of course, these stats change from country to country and this makes it hard to launch a website with a single audience in mind. For universities, this becomes incredibly important in terms of international students recruitment.

Similarly shocking, we have learnt that the combination of Safari and Chrome, the major mobile browsers reach 93% on WiFi networks but only 88% on 3G networks; this suggests that connections speeds still matter to people, who might opt for different, more traffic-efficient browsers in connectivity-challenged environments (for example, OperaMini goes up from 1% to 4%)

One good practical piece of advice is to adopt the RAIL Approach, promoted by Google, which is a user-centric performance model that takes into consideration four aspects of performance: response, animation, idle time and loading. The combination of these aspects, each of which has its own ‘maximum allowed time’ before the user gets frustrated or abandons the activity, requires a delicate balance.

There was also some good level of discussion around the very popular “responsive web design”, a technique that has become a goal in itself. The speaker suggested that this should be just a tool, rather than a goal: users don’t care about “responsive”, they care about “fast”. Never forget the users is a good motto for everyone working in IT.

Summary
Velocity’s first day has been a very hands on day. The overall take-home lesson is simple: managing performance requires some sound science, but with adequate tools and resources it’s not impossible to do it on a shoestring budget and in an effective way. As I’m an advocate of internal resource control and management with respect to outsourcing, today’s talks have surely provided me with some great insight on how to achieve this smartly.

Aside from this summary, I’ve also been taking some technical notes, which are available here and will also contain notes from the future sessions.