Mouse in the House

Posted by mark on February 17th, 2011

Ever wonder how fast your website (or any other website) loads from different locations around the world? Especially if your site relies in part on third party content, the user experience at various cities can be very different indeed!

Using our WatchMouse Performance Monitoring service API, Loads.in is a convenient webmaster tool that allows you to measure just how fast a website loads in a real browser from over 50 locations worldwide – on every continent except Antarctica!

Websites can be particularly susceptible to slow page load speeds when they need to load a high amount of components (images, JavaScript, third party content) to render the complete page.

The free, Loads.in tool checks your site utilizing a real browser, and provides snapshots and waterfall charts for each check.  A selection of browser profiles is available too, and include Safari, Chrome, Internet Explorer and Firefox

Simply enter the full URL of the page you want to check in Loads.in, and the page is retrieved by a browser at a random location. For each subsequent check you can choose a specific location*.

The Loads.in results presented include:

• The page load time of the website
• Snapshots at different times during the loading of the page
• Errors or warnings if they occur
• A complete timing breakdown of all elements of your page in a “waterfall chart”
• The option to download the timing results in the HTTP Archive (HAR) format

*Locations include: Amsterdam, Antwerp, Cologne, Copenhagen, Dublin, Glasgow, Groningen, Lille, Lisbon, London, Madrid, Manchester, Munich, Oslo, Padova, Paris, Stockholm, Zurich, Bucharest, Kharkov (Ukraine), Krakow, Moscow, Vilnius (Lithuania), Melbourne, Sydney, Cape Town, Bangkok, Haifa (Israel), Jakarta, Kuala Lampur, Mumbai, Nagano, Shanghai, Singapore, Guadalajara, Vancouver, Austin, Chicago, Dallas, Florida and NYC.

Check it out and let us know what you think. We value your feedback and hope you find Loads.In to be a useful tool and resource.

Happy Monitoring,

Mark Pors
CTO & co-founder

Posted by mark on February 11th, 2011

Monitor metrics vs. User satisfaction

Monitoring the performance of your website pages generates lots of interesting metrics like resolve time, connect time, time to first byte, DOM-ready, first visual (in real browser monitors) and many more.

These metrics:

• provide insight in what might cause a page being slow,
• shows page load time trends over time,
• can trigger alerts in case of a temporary performance drops,
• show the effect of request originating from different geographical locations, etc, etc.

What these numbers do not say is how this affects the user satisfaction of  visitors of your website. This is where the Apdex comes in.

The Application Performance Index (Apdex)

The purpose of the Apdex score is to convert performance metrics into insights about user satisfaction. It allows you to specify a threshold that indicates whether your service is operating satisfactory or not.

See the Apdex website for further information and exact definitions.

WatchMouse and the Apdex score

Next to the WatchMouse Site Performance Index, which combines page load time and downtime in a single metric, we recently introduced the Apdex as a new chart type. Both metrics are also available in the custom PDF reports.

The charts below show the Apdex for one page over time:

The charts can also be used to compare Apdex scores between pages:

These charts can now be interpreted as showing the fraction of satisfied visitors on your site: 0 = no users satisfied, 1 = all users satisfied

Note: To make sure that the Apdex chart is based on valid assumptions, the monitor or monitors you use should fulfil the following condition:

Time-out > First limit * 4

The time-out and first limit can be set in the expert mode of you monitor settings. By default the settings will be Apdex compliant, and a warning will be shown if the settings are not Apdex compliant.

What’s next?

The Apdex charts in the WatchMouse reporting are just a first step in reporting customer satisfaction in terms of performance. Now what I would like to know…

• Are the new Apdex charts useful for you?
• How will you use them in your organisation?
• What else would you like to see related to Apdex?

Please let us know in the comments!

Mark Pors
CTO & co-founder

Posted by mark on February 7th, 2011

Public Status Health Dashboard 4.0 released

Over the weekend we had a major release of our Public Status Pages. I’m very exited about the improvements both on the back-end and in functionality for our customers.

In this article I’d like to walk you through the improvements and invite you to share your suggestions for the next release.

The WatchMouse Public Status Pages

For those of you not familiar with our Public Status Pages yet, I included a short summary on the what, why, and who.

What is a Public Status Page?

A WatchMouse Public Status Page enables your organisation to display information about the availability and performance of your critical services. You can post announcements, annotate current issues, and optionally set up a special host name (CNAME) so people can access the status page on your domain, e.g. status.yourdomain.com. It is an easy control channel through which you can transparently inform visitors about the status of your sites and web services.

All WatchMouse Public Status Pages are hosted on Amazon’s Cloud infrastructure so they are available even if your site or service is not. Read more here.

Why Public Status Pages are important.

The single most important reason to have a Public Status Page or Health Dashboard is to have communication channels in place well before a ‘crisis’ strikes. Find more about why you need a status page in another article on this blog “Transparency is Critical When Sites #FAIL“.

Who is using Public Status Pages?

Here is a list of some of our more well known customers using the WatchMouse Public Status Pages:

• Twitter - http://dev.twitter.com/status
• Mozilla - http://status.mozilla.com/
• Wikipedia - http://status.wikimedia.org/
• WordPress - http://status.automattic.com/
• Read Write Web - http://status.readwriteweb.com/
• Bit.ly - http://status.bit.ly/
• and many more!

More status dashboards (powered by WatchMouse and others too) can be found here.

Improvements in release 4.0

So what is improved in this new release?

• New powerful architecture and storage engines, based on MongoDB
• Highly available and even more scalable (still hosted in the AWS cloud)
• Always up-to-date with latest check results, instead of updated on changes in monitor status
• ‘Moving’ uptime figures over last 24h instead of today’s uptime
• Better “per country” indication, now averages over the last N checks
• Interactive charts, powered by the Google Visualization API
• Zoom-able world map for more details in Europe
• Clear daily uptime charts
• Improved console
• HTML support for public notes, including an HTML editor
• SSL support

What’s next? Your opinion counts!

Some ideas we already have and working on for the next release of the WatchMouse Public Status Pages are:

• Easier (self service) customization directly through the Public Status Pages console
• Browsing back in time (for all charts and history section). The back-end system is ready, now working on the front end
• Long term (monthly) charts
• Private Status Pages (only accessible by authenticated users)
• Real Time Status Pages (Comet/WebSockets support)
• Public Status Widgets for easy integration into many popular blogging engines.

So what would you like to see in our next release? Please let us know in the comments, or contact us by creating a helpdesk ticket.

Mark Pors
CTO & co-founder

Posted by stan on December 28th, 2010

We monitored and tested the leading shopping websites in six different countries to see how they fared in the lead up to and during the holiday shopping weeks.

The 300-plus websites we tested in the United States, United Kingdom, Spain, Germany, Belgium, and The Netherlands performed quite well overall. The country with the most sites in the 100% uptime category was Germany with 33% of the 30 tested sites experiencing no downtime during the reporting period. Coming in second was the United Kingdom with 31% of the 89 websites experiencing 100% uptime, followed by Spain with 29% of 17 websites and the United States with 27% of 100 sites with no downtime. The Netherlands had only 12% of the 91 sites tested with 100% uptime, while Belgium sites performed worst of all with only 10% of the 83 sites tested with 100% uptime.

You can read the full performance reports and view a list of the websites monitored in each country by clicking on the following: United States, United Kingdom, Spain, Germany, Belgium, and The Netherlands. You can also view the current live health of each of the 327 websites we monitored, by visiting the Public Status Pages for each country: United States, United Kingdom, Spain, Germany, Belgium, and The Netherlands.

Happy Holidays and we look forward to sharing more monitoring news in 2011!

The WatchMouse Team

Posted by mark on December 17th, 2010

The URL shortener tr.im appears to be near death. They pulled the plug quite a while ago and are no longer accepting URL shortening requests, but they kept their systems for existing shortened URL’s running.

This past April, the company stated “we would like to shut down the API and redirection service by the end of 2010″. The last few days the service shows a downtime of eight hours per day – perhaps a signal of their last flickerings of life?

WatchMouse monitors 24 URL shorteners, and currently four of them are broken:

• snurl.com: this service has an average uptime of under 90%, rendering the service useless.
• to.: down and out of business
• tr.im: shuttering and out of business
• twurl.cc: down and out of business

Some companies that cease operating their URL shortener leave the shortened links live even though they are in essence out of business; others do not.

Many links on the Internet break because companies providing URL shorteners go out of business or change their policies or priorities. The people that have used these services to shorten a URL are typically not given advance warning, and are simply left with error messages.

And then there is Digg. They had a URL shortener service, but decided to use it internally only. According to this article the promise was made, however, to keep existing short URLs working, but our monitoring reveals that is no longer the case as of November 30.

We recommend you choose your URL shortener wisely, e.g. bit.ly whose core business is URL shortening and should be around for a long time. (Disclaimer: bit.ly is a WatchMouse customer).

Happy Holidays and Happy URL Shortening!
The WatchMouse Team

Posted by mark on November 23rd, 2010

Cloud computing has made it easy to build applications that run reliably even under a heavy load, and developers need to know if and when the cloud, and thus their application, is having problems.
We’re very pleased to announce today that we’ve acquired Cloud Status for iPhone, an application originally created by Alasdair Allan, noted author, software programmer and expert iOS developer. Our collaboration and acquisition of the Cloud Status for iPhone app has allowed us to not only add new features in the latest version 4.4 release, but also to make the app available for FREE to the developer community and IT departments around the world who depend on cloud based services to run their businesses.

The Cloud Status for iPhone version 4.4 release includes the following features:

• FREE to download
• Support for iOS4
• Support and reporting for Amazon Web Services, Google App Engine, Google Apps, Microsoft Windows Azure, and Rackspace Cloud
• Fully supported retina display in iPhone 4

Each of the supported cloud services has a separate page and details the status for the various services provided. A quick read indicator denotes the status for each service: the status for the service is good, there is a problem with the service, or the service is down. Clicking on each service component provides further information as to the current status of that component, and any problems it might be experiencing.

We greatly respect and admire the work of Alasdair Allen on the iOS platform, and we plan to work with him in the future to create additional applications that will support other WatchMouse performance monitoring services.

For more information click here.

Posted by stan on November 8th, 2010

We’re pleased to announce a new addition to our monitoring services today (drum roll, please) – Real Browser Monitoring!

What is Real Browser Monitoring?

Real Browser Monitoring accurately measures the performance of real browser visits. By pinpointing performance issues using an actual browser, this service enables businesses to verify that their site and web applications are delivering the best possible user experience.

Traditional site monitoring tools emulate browsers to check a site’s availability and performance. That is perfectly suitable to get the performance and uptime under ideal conditions. Real Browser Monitoring however, allows you to verify a site’s performance as perceived by the end-user using a real browser.

Why should you care?

Today’s sites commonly use Web 2.0 technologies (Ajax, JavaScript, CSS and Flash) to allow collaboration between the components of a site from their different global locations. Sure, your site may load perfectly and swiftly from the office, and from other locations in your area, but how about other cities or other countries? If your site attracts more than a regional community of users, you should be thinking about using real browsers to monitor your website’s speed. Your own content may be optimized locally, but are all of your other components optimal as well? And, how optimized is your website for the international visitors you wish to reach?

The actual performance of your website (as your visitors experience it) is affected by:

• Third-party content (Google ads, Facebook applets, Twitter feeds, Discus forums, widgets, etc.)
• Dynamic content executed in the browser (JavaScript, AJAX, CSS)
• Effective use of a content distribution network
• The (network) distance between your site and your visitors, i.e. the UK, France, and Germany are close, but Australia and Brazil may be far away

Real Browser Monitoring checks that each component of a Web 2.0 site is correctly responding, and functionally verifies that your composite sites or sites are working – from the web server right through to the end-user’s browser. It benchmarks a website, reports the true end-user’s experience, and offers insight above and beyond regular website monitoring tools that synthetically emulate browsers to check a site’s availability and performance

Along with regular monitoring, we advise that you monitor full page load times with a real browser on an ongoing basis, and from all the regions and countries where your (potential) customers may reside.

The only way to accurately monitor the performance of each component of a Web 2.0 site is to use a real browser to execute the JavaScript.

You can start right now!

Already a WatchMouse client? If you are subscribed to one of our current plans you’re al set to use this new type of monitoring: Each of the plans included ‘scripting monitors’ which can now be used for either scripts or for real browser monitors. And you can extend your plan with extra real browser monitors here. Your first time? Then sign up for a free 30-day trial to experience real browser monitoring yourself right now!

Feel free to contact us if you have questions.

Happy Monitoring!

The WatchMouse Team

More on Real Browser Monitoring…

Posted by stan on October 31st, 2010

When Gmail is slow, Amazon trips, when there is a Facebook issue, or Foursquare’s API crashes, people get upset, and tens of thousands reveal their anxiety on social networks.

Every time there is a major outage, security issue, or malfunction we see this pattern of raised anxiety, doubts, and questioning of services in the cloud in general.

This makes perfect sense of course, as Web apps and Web services have become more prevalent and are now an essential part of our daily habits and work. The advantage of Web apps is compelling: simply grab a PC, iPad, laptop or mobile phone and you have everything at your fingertips. But there is a downside too: we become dependent on the cloud infrastructure and our connectivity to the network.

The reality is that software and services break. Desktop software normally breaks with one user at a time, although millions will be affected over time. If a Web app breaks however, hundreds of thousands are affected at the same time. There is also a psychological effect: when software on a computer fails, people often feel (partly) responsible for it. With Web apps, the provider is the only one to blame.

So even if Web apps are far more reliable than local apps – and I believe this is often the case – the public outcry is far more extensive. Especially now, when typing “#fail” on social networks like Twitter and Facebook, is only seconds away.

What does this imply for the companies behind the Web apps we use every day? The single most important thing is to have the communication channels in place well before a ‘crisis’ strikes.

I personally suggest having the following at minimum:

1. A blog or status page that is hosted independent (and scalable!) from the main website and services. Easy, predictable, standard names should be used: blog.company-or-brand.comand status.company-or-brand.com, respectively. A good example: status.readwriteweb.com. Preferably these pages should include up-to-date, live, stats.
2. A Twitter name where one can post quick updates. If possible @company-or-brand should be used here too. Example: @rww.

This first level of transparency makes it easy for people to get informed and immediately results in lower anxiety levels. This in turn, helps to stamp out rogue stories in times of crisis, and reduces the load on the company’s customer service contact center.

Many companies have set up public status pages already.

• Admit failure as soon as possible, preferably by someone high up in the organization
• Make sure the posts and updates sound human, no standard sound bites
• Explain in detail who and what is affected (which regions, percentage of customers, what services, etc.)
• Publish a detailed timeline of the outage, and start maintaining this immediately after the first event
• Share detailed post mortem reports and lessons learned after the crisis is over
• Read more here for a more detailed analysis of the psychology of transparency

If these guidelines are followed, the added benefit is that it actually induces and instills a higher trust in the company and its brand – not less. It also gets the message across quickly and efficiently, so it can then be relayed across social networks, instead of leaving it up to the guesses of the public or the media. Finally, it will save serious money in the company’s contact center, as it sets the right expectations.

Companies that are transparent about issues regarding their services will actually gain kudos and trust.

So the next time there is an issue with your favorite application on the Web and www.your-favorite-app.com is not working: you might want to check out status.your-favorite-app.com to see if there is up-to-date information before you type “#fail”. A Public Status Page may just be waiting for you there.

Stan P. van de Burgt
CEO and co-founder of WatchMouse

Posted by stan on September 23rd, 2010

Hola! We’re excited to announce that WatchMouse was named one of 10 Young Bull winners at the inaugural 2010 Bully Award, honoring the leading technology, media and telecommunications (TMT) companies in Europe as presented by White Bull Summits.

A total of 60 European companies were named as finalists for the 2010 Bully Award, and 30 were honored as winners. The deserving winners were selected from a pool of entries that included hundreds of nominated European TMT companies.

The winners fall into three categories:

• Yearlings: firms that seek or have received angel/seed rounds or equivalent; classic start-ups
• Young Bulls: firms that seek or have received Series A financing; early stage companies
• Longhorns: post Series A firms; growth stage companies

Here’s what Farley Duvall, Founder and Chairman of White Bull Summits had to say:

“The Bully Award winners were selected for innovation and values of excellence in the TMT sector. Each of the 2010 Bully Award winners have been recognized as a leader in their field, with a bright business proposition and meaningful market strategies, driven by a rich understanding of customer needs and technological solutions. We searched for the ‘wily beast’ we believe lives inside all successful businesses, and very carefully reviewed start-ups, early-stage and growth-stage companies. We believe the 30 winners we selected have demonstrated clear potential for further growth, including a high probability for exit in the next few years.”

Winning the White Bull ‘Young Bull’ award is significant for us for two reasons: it is recognition and further validation of our product, and it also speaks to our ability to challenge the old bulls in the monitoring space. The award is also a notable milestone for us after posting a remarkable 60% growth year-on-year in 2010.

We’re very honored to win the Bully Award, and to be recognized as one of the leading TMT companies in Europe. It’s extremely validating to be named a winner of this prestigious award, and we accept and share it with the entire WatchMouse team, community and our clients.

For a complete list of the 2010 White Bull Bully Award winners, please visit www.whitebull.com/latest.

Adios amigos and we’ll be back soon with more WatchMouse news!

Posted by stan on August 29th, 2010

We’re excited to announce that WatchMouse has been named as a finalist for the inaugural White Bull Summits 2010 Bully Award, honoring the leading technology, media and telecommunications (TMT) companies in Europe.

A total of sixty European companies were named as finalists for the 2010 Bully Award, and were selected from hundreds of entries by nominated European TMT companies. You can see a complete list of the 2010 White Bull Bully Award finalists here: www.whitebull.com/latest.

Here’s what the founders of White Bull Summits had to say:

“The Bully Award finalists were selected first and foremost for values of excellence and innovation in the TMT sector,” stated Farley Duvall, Founder and Chairman of White Bull Summits. “WatchMouse has been recognized as a leader in its field, with a bright business proposition and meaningful market strategies, driven by a rich understanding of customer needs and technological solutions. We searched for the ‘wily beast’ we believe lives inside all successful businesses. We carefully reviewed seed, angel, early-stage and growth-stage companies and have chosen those firms as finalists that we believe have demonstrated clear potential for further growth, and are particularly deserving of a moment in the spotlight.”

We’re very honored to be named a Bully Award finalist, and to be recognized as one of the leading TMT companies in Europe. It’s extremely validating to be named as a finalist for this prestigious award, and we accept and share it with the entire WatchMouse team.

The thirty winners will be announced live at the White Bull Summit 2010 in Barcelona, September 20 – 22, and we’re hoping to be among them.

Stay tuned!