If so I’d love to get your feedback on what was good & bad about the event. I could have handed out feedback sheets at the end of the event… but everyone would have been too busy eating and drinking to fill them out, anyway!

So if you can please add your comments to this blog post or email me directly at info at seriticonsulting.com that would be very much appreciated.

If you could think about things like:

1. Format of the day and how sessions were allocated etc
2. Content of the sessions
3. Venue
4. Food&Drink
5. Swag & Sponsorship
6. Audio/Visual quality
7. Communication before, during and after the event
8. Overall atmostphere etc

But don’t feel you have to limit yourself to these topics – all idea and suggestions welcome!

You can also leave comments over on the London Web Performance Facebook page if you want! https://www.facebook.com/pages/London-Web-Performance-Meetup/177580535617183?fref=ts

“#WebPerfDays London was awesome. Tremendous job by @TheOpsMgr& crew. Thx #Facebook, #Dyn, & other sponsors. Love the longer conversations.” – @souders, Steve Souders, Google.com

#WebPerfDays was awesome  

130+ enthusiastic WebPerf people sharing ideas and knowledge in an unConference format, with lots of jokes, laughter, food and drink along the way!

This post is just a quick summary of the slide decks and other material that was referenced during the day but they best parts were not captured in PowerPoint or Keynote but in the many lively conversations going on all around the amazing Facebook venue!

Don’t forget to join the London Web Performance Meetup and be a part of the monthly Meetups! We are always looking for speakers, venues and sponsors to keep the #WebPerfDays and #VelocityConf message alive!

The group was started in January 2011 after being inspired by the work Sergey Chernyshev was doing in New York.

Our group has now grown to have 422 members and continues to get excellent ratings from those who attend (4.61 out of 5 for the last meetup at Betfair!) and the Web Performance Meetup movement has expanded to include San Francisco, Boston, Los Angeles, Paris, Vienna, Atlanta, Austin, San Diego, Herzeliyya (Israel) , Amsterdam, Berlin, Hamburg and Cologne.

Our group has members from some of the UK’s largest online organisations such as sky.com, Betfair.com, totaljobs.com, Lonelyplanet.com, UBM, LBi, Telegraph, FT.com, Guardian, IPC, Playfish.com, yell.com etc etc. Everyone has been very open and willing to both learn from others and share their own experiences and this has been a huge factor in the group’s success.

Partly because of the success of the #WebPerf meetup movement the Velocity Conference has come to Europe and some of us were even lucky to speak at it!  O’Reilly have been great supporters of the London Web Performance Meetup group through donating prizes and free conference passes to help attract members. Our sincere thanks to the team there, particularly JWB!

We’ve had some excellent speakers at @LDNWebPerf who have also presented at Velocity and other major events such as Joshua Bixby (Strangeloop), Andreas Grabner (Dynatrace) and Jyoti Bansal (AppDynamics) plus “local heroes” like Perry Dyball & Ged Waring (Seatwave.com), Andrew Betts (Assanka) and Mick McGuinness (ApplicationPerformance.com).

We’ve got a great schedule already taking shape for 2012 and we look forward to more talks, hopefully more sponsors, and lots of great post-presentation conversations over a pint or two.

For those who weren’t able to make it to some of the previous meetups I have listed the slide decks, where possible, below!

January - Web Performance Case Study - Seatwave.com

February - Web Performance Optimisation 101

March - Performance Automation - faster ways to do make your website faster

April - Managing Performance in the Cloud - Jyoti Bansal, CEO of AppDynamics

[cant find the slides, will have to get back to you on that…]

May - Case Studies: Performance for "mortal companies" - Josh Bixby

July - Velocity 2011 Feedback session

August - Building a high performance HTML5 Application - the new FT.com HTML5 web app

September - Measuring Mobile Performance

October - HTML5 WebSockets and the Realtime Web

November - Speed War (Operation "Jungle Cheetah") - WPO @ The Times

December - Xmas Special - Web Performance at Betfair

I was at my London Web Performance Meetup the other night at Betfair and Andrew Harding, a perf QA engineer there, gave a talk on “Continuous Integration - A Performance Engineer’s Tale” in which he talked about how they were tackling this challenge at Betfair.

He said something that resonated with something that I have been thinking about for a while now, to whit,

“We’ve had to look at separating load injection from performance measurement”.

Now, Betfair did this for various reasons, mainly to do with the need to keep the perf testing environment “warm” – it takes too long to warm up the environment, load the caches, overcome TCP “slow start” etc and achieve a “stable” system – compared to the time they might have between check-ins/builds etc. So it’s easier to keep a constant load flowing through the environment to keep as much of the system as possible “warmed up” until you deploy your incremental changes in the next build and do your testing.

Because they use “traditional” load testing tools like LoadRunner which are designed to “start script – inject load – measure response – stop script – generate report” this causes a problem. They aren’t “starting and stopping” all the time, hence they don’t get the nice reports out the end. [I might be over-simplifying this, I am not a LoadRunner expert, but bear with me for now!].

So they “measure performance” using other tools e.g. Real-user monitoring (RUM) or Application Performance Management (APM) tools outside of the load-injection toolset.

Now, I see this as a real problem for the vendor’s like HP who make LoadRunner. LoadRunner has lots of awesome features but if you are just using it to “generate load” and not using its measurement and reporting features then there are far cheaper (and open-source) alternatives. The rash of “cloud-based” load testing services that are springing up off the back of Amazon EC2 are also increasing the downward pressure on the costs of the “traditional” vendor solutions.

The “cloud testing” vendors raise another issue with the measurement of performance “from the point of injection”.

Measurement requires, as much as possible, a stable platform from which to measure, which in it’s current form the cloud most certainly is not. Study after study after study shows that performance in the cloud is variable, and even more so under load. Some people have even given up on the cloud for that reason and moved back to dedicated servers (http://code.mixpanel.com/2011/10/27/why-we-moved-off-the-cloud/. Well worth reading the comments below as a lot of people both agree and disagree).

So if you want timing resolution to within (at least) 10msec to measure variation in time-to-first-byte then doing it using a varying “measuring stick” isn’t going to work. So, another argument for separating load injection from performance measurement.

But this isn’t the fundamental reason I am concerned that things might be “broken” for the current load-testing model.

Currently load-testing is [mostly] based on a classic HTTP request/response paradigm – “start a timer, make a HTTP request, stop the timer, record the duration, repeat as necessary, report the outcome”.

Ok, so @LDNWebPerf we’ve looked at HTML5 apps and at WebSockets and, guess what, the web apps of the future can (and do) break this simplistic “classic HTTP request/response paradigm”.

HTML5 apps might make extensive use of localStorage to eliminate roundtrips and WebSockets changes everything into ONE HTTP request followed by a lot of BI-DIRECTIONAL communication over the WebSocket connection. The same issue exists with some of the other “push” techniques that might use chunked-encoding to send data asynchronously down a “long-lived” request/response. SPDY might cause problems too because of the way it multi-plexes the Request/Responses too!

So how are you going to measure your “performance under load” when there isn’t a nice “Request/Response” to “start/stop” your timer?

Well, the answer is… I don’t know, yet.

There are two immediately “obvious” solutions – go “up the stack” or “down the stack”.

By “up the stack” I mean that the load testing tool much be much more “web application aware” – it must be inside the browser, seeing exactly how the web application is sending/receiving information especially over WebSockets and be “instrumented” to measure it.

It might also require it to be “framework-aware” so that it understands frameworks like Dojo/Comet/Jquery/whatever and “knows” how the methods they use to send/receive information and how to inject instrumentation into those frameworks to measure them.

By “down the stack” I mean back to network sniffing and measurement “on the wire” to see exactly what how long things take, but again probably with the requirement to be more “application-aware” so it can re-assemble the network packets and translate them into HTTP, WebSocket and application-level performance timing data that’s more meaningful to the performance engineer and the application developer.

Anyway, I am not sure that I have explained this as articulately as I would have liked but my basic message is that:

1. If we can’t measure performance from the “point of injection”, for whatever reason, then you will need to invest in other tools that can measure performance from different locations (e.g. RUM, APM etc)
2. Hence, if all you are doing is generating load, then you’ll rapidly become a commodity item and be paid accordingly…
3. New web technologies like WebSockets can change the HTTP Request/Response paradigm making the job of performance measurement more difficult so the tools will need to evolve and perhaps become more “application-aware”.

I’d love to hear what everyone else thinks so please comment away!

The aims of the WPA are simple – to provide UK e-businesses with the information they need to make their websites FAST!

The new WPA website has a knowledgebase of Web Performance Optimisation (WPO) information that will be continually updated with the latest information from WPA members and the leading WPO experts from around the world. www.websiteperformance.org.uk/resources

If you need help in making your website fast then the WPA members are here to help! www.websiteperformance.org.uk/services.

If you want to get involved or contribute some material to the WPA knowledgebase please let me know.

All the presentations were well received and there was an excellent turnout from some of the UK’s top e-commerce companies.

The event was also the launch of the UK Web Performance Alliance www.websiteperformance.org.uk – a new industry collective to help promote performance in the UK. More about the WPA in a later post!

My slides from the keynote presentation are below!

It covers the basics about measuring mobile performance, and what techniques are appropriate for your needs.

GIVE THE PUBLIC THE CRIMINAL JUSTICE SYSTEM THEY WANT

The country is in danger of fighting violence, arson and looting on our city streets with
sound-bites.

The Metropolitan Police Federation – and, we suspect, every victim of the recent riots –
want to see these criminal acts fought, instead, with a resilient, efficient, properly resourced
criminal justice system.

And this is something which, at the moment, the country does not have.

Prime Minister David Cameron has been quoted as saying: "Whatever resources the police
need, they will get."

Yet police manpower resources are being diminished. Her Majesty's Inspectorate of
Constabulary Report 'Adapting to Austerity' says that, as a result of the cut to its budget, the
Police Service in England and Wales is reducing in size by 16,200 police officers.

This is almost exactly the number of officers that was deemed necessary be on the streets
to protect London from criminal attack on the night of August 9.

Mr Cameron was also quoted as saying he had asked the police if they had everything they
needed and had been told they had.

Which police were these, precisely? Senior officers? The very epitome of back office people
which the Government has vowed to eradicate?

Or front-line police such as the frighteningly outnumbered officers attempting against the
odds to protect the streets of Tottenham during the first night of rioting?

Mr Cameron says the police will be allowed to use baton rounds – "Whatever tactics the
police feel they need to employ, they will have legal backing to do so."

So what? The police have legal backing to use ordinary batons, too, but many fear to do so
in case one strike goes wrong and the officer involved faces a criminal charge.What the police need is not meaningless words but cast-iron assurances at the highest political level that they will be allowed to confront arsonists and looters effectively without fear of themselves ending up in prison.

The Government has made much of its intention to reduce police bureaucracy. What the
police need here is action, not words.

As things stand, a dog handler is being obliged to waste time writing a statement when their
dog has, on instruction, barked at someone – because this is deemed to be a use of force.
Similarly, an officer who draws a baton in the face of a hostile crowd must record their
actions in detail.

Let's get rid of this bureaucracy, not talk about getting rid of it.

From the front line officers' point of view, London's police are already being squeezed until
the pips squeak.

In the Met, leave has already been cancelled for the next week; a week later there is no
leave due to the Notting Hill Carnival and there is no leave the weekend after that due to a
planned English Defence League march.

Officers will continue to do everything they can to protect London and its citizens; however
they cannot sustain extended hours of working indefinitely.

And while they endure gruelling duties which last up to 20 straight hours at a stretch, they
know that shortly their payment for doing so is to be slashed.

Burning buildings and looted premises provided dramatic pictures for media covering the
riots. But from the point of view of police, the most striking images were those of youths
and children committing acts of destruction and theft without even bothering to cover their
faces in front of the television cameras.

Why? Because they do not fear the criminal justice system. And that's not surprising.

Many of the young people arrested by Metropolitan officers during the riots were, on
Crown Prosecution Service advice, given police bail, meaning that they were immediately
back on the streets again. Some will doubtless vanish before any further action can be taken
against them.

The Prime Minister has stated that anyone convicted of violent disorder will be sent to
prison.We will wait and see whether the CPS and courts share his appetite for incarceration. In
view of the thousands of breached community sentences and ASBOs which litter the
Criminal Justice System – not to mention the millions of pounds in unpaid fines – we have
little faith that they do.
What the police need is a more resolute approach in punishing offenders, who are all too
often are freed by the courts to re-offend, obliging officers to expend time, energy and
resources in re-arresting them for new offences time and time again.

Metropolitan Police officers have total sympathy with, and care about, those people who
have lost their homes and possessions or their businesses, which they have spent a lifetime
building up. Officers are committed to doing all they can to prevent fresh victimisation by
the mob - or indeed, any criminal.

But those in power must listen to the people who know most about policing the streets –
the frontline officers who actually do it – and must mean it when they say these men and
women will be properly resourced.

And the bigger machine of which the police are just a part – the criminal justice machine –
must be made to work in the way the public want it to work and have a right to expect it to
work.

John Tully
Vice Chairman
Metropolitan Police Federation
For further information, please contact John Tully on 07825 169648

(original PDF is here - http://www.metfed.org.uk/support/uploads/1313053433Statement%20by%20Metropolitan%20Police%20Federation%20Vice%20Chairman%20John%20Tully.pdf)

It is a pure scam to get you to install remote control software on your PC and to then use your PC for malicious purpose e.g. what is called a “botnet” of infected PC’s that can be used for spamming, denial of service or other illegal activities.

The woman who called me was reading from a script and offered “step by step instructions to fix my PC”.

The first step is to take you to your “Computer Management” console and then to “Event Viewer” and ask you to look for “all the red crosses that are error messages showing you that your PC is infected/corrupted”.

This is to create fear and doubt in the PC owner.

In reality, just about every PC will have “red error messages” that are utterly harmless and can safely be ignored.

The next step is to direct you to this website – www.ammyy.com – and to try and get you to install their remote control software.

DO NOT DO IT!

I REPEAT…

DO NOT INSTALL ANY SOFTWARE

FROM ANYONE THAT RINGS YOU

UNSOLICITED!

THIS IS A SCAM!

My advice at this point is to (1) hang up the phone or (2) verbal abuse the lying filth at the other end and then hang up the phone.

Step (3) is to warn all your friends and relatives, particularly the less computer savvy amongst them and make sure that they don’t fall prey to the scammers.

If you have been scammed and installed the software the only (and I do mean ONLY) safe course is to wipe your hard drive, re-format it and re-install your operating system from scratch.

There is some advice on the internet about using Control Panel to un-install the software from Add/Remove Programmes but personally I wouldn’t take the risk – once your PC has been seriously infected you cannot rely on Add/Remove programmes, and since you probably over-rode your anti-virus programme during the installation of the malicious software I wouldn’t trust it, either.

So it’s back to the installation DVD and installing it from scratch.

Consider it your penance  

If you want a laugh, watch this great video from a young lady that almost, almost fell for the scam!

Who should be on your WPO team? but today I want to explore more about the economics of web performance and why I think the future of WPO/WCO/#WebPerf is automation.

My strongly held view is that WPO/WCO isn’t a project, it’s an on-going process, which implies that if you are going to form your own in-house WPO team you are looking at a significant investment in time, human resources and money.

If you had an 8 person team as per my earlier post, at a base salary here in the UK of about £40K that’s £320K in basic salaries alone. Add in pensions, training, office space, PC’s etc lets add 25% for extras on top, giving us £400K (minimum).

If we cost that out over the 3 years we normally use for software purchase lifecycles, that’s £1.2M.

Ouch.

Ok, let’s dial that right back and say… that you only hire 1 permanent employee to be the WPO guru and do the rest of the project with contractors over 3 months (60 days).

£40K base + 25% for extras for the permie times 3 years = £150K, 7 contractors on £400/day times 60 days = £168K. Grand total is now £318K.

Much more palatable!

OK… so let’s say I look at an automated solution instead.

The prices vary for these depending on whether you use an appliance, per server or SaaS solution but for the sake of argument lets say I chose a per server solution and it’s going to cost me £8K per web server per year.

If I have 10 servers that’s 10 servers times £8K times 3 years = £240K.

I’m £78K in front (£240K vs £318K), and haven’t had to recruit any staff, pay huge recruitment fees, find office space etc etc.

Plus with an automated solution you are likely to get your site running faster, sooner, without the uncertainties and delays of a software project, and hence start to reap the conversion benefits of the faster site immediately.

Unrealistic?

Well, I was playing around with an automated solution this week and with less that a day’s effort I got these results for a consumer e-commerce site.

Before – 215 requests, 7.7 seconds

After – 71 requests, 4.7 seconds

In a full scale implementation the accelerated site performance would be even faster because this was performed as a quick proof of concept using a micro EC2 instance as the acceleration solution host to reverse proxy the origin server. In a real implementation it would be optimised/transformed “at origin” thus eliminating the additional round trip delays to the Amazon data centre in Ireland.

Anyway. my point is that unless you run very large numbers of servers or have lots of IT staff with web perfomance skills lying around doing nothing the economics of web performance optimisation heavily favour automated solutions, even if you use commercial product as I have discussed in this post.

If you throw in open source solutions like Site Speed the economics favour automation even more.

So it’s worth doing the numbers before committing to a complex development project if there are some automated “quick wins” just waiting to be grabbed!"