Monitor the holistic performance of your website infrastructure with Dotcom-Monitor, a complete cloud-based website monitoring solution. Analyze everything from the underlying hardware to the response time of individual page elements over time, receiving alerts the moment issues are detected. Utilize integrated diagnostic tools to improve website speed and optimize your infrastructure.
With Dotcom-Monitor, the website monitoring process is simple:
There are two basic metrics often used to track the performance of a website. The first is page load speed – which records the amount of time that passes from the initial request to the server to the point that the server sends a FIN packet to close the connection.
The second metric is time to interact – which measures how long a user must wait before they can begin reading or interacting with content on the web page. The first website monitoring metric, page load speed, is the easiest to measure, as closing the connection is a discrete task. Time to interact, however, can be more nebulous depending upon the content of the website. Time to interact is relevant when you are concerned with holding a user's attention and decreasing the user bounce rate of a website. As page load speed is much easier to measure via website monitoring, and typically provides more consistent results from one test to the next, most companies have focused on monitoring the page load speed metric as a baseline for website performance.
With simple headless HTTP website monitoring you can setup individual key page monitors – ie: home pages, marketing landing pages, and key conversion pages. Start with something as simple as response code monitoring, and receive an alert if the page returns any code other than a 200 OK status. When returning a 404 not found, 500 server error, etc – alerts are immediately triggered indicating the negative response code. After HTTP website monitoring is initialized, users may also opt to implement more robust monitoring methods to ensure that every single element on the monitored pages are loading properly. This is called full page website monitoring, and is part of Dotcom-Monitor's BrowserView platform.
As you gather website monitoring data you begin to build a story. This story can easily be seen in a line chart as the responsiveness of the website over time. Spikes and plateaus can be quick indicators that something changed in the system that negatively affected page load speed. Using the available waterfall charts you can easily drill down into these peaks and valleys to identify individual elements that may be the cause of the changes in website response times. Clicking on the element trend of an individual image, script or file will show you how the download times of that element have changed over time. This is particularly useful when communicating between departments such as WebOps and Marketing, as IT can show marketing how the content they produce is directly affecting the user experience on the website.
Dotcom-Monitor maintains website monitoring servers in over two dozen locations around the world. These locations have been strategically selected to provide coverage from the most important geographic areas on six continents.
Servers behind the Great Firewall of China monitor website application performance from an inside-out perspective, giving insight into exactly how Chinese users are viewing your site, and how the Great Firewall of China (GFW) may be impacting your website's performance. As the Chinese economy continues to grow, more and more web-based companies are concerned with reaching the Chinese populous. With stringent control over the flow of information by the Chinese government, it is important to monitor the accessibility and page load speed of websites from behind the Great Firewall of China. Dotcom-Monitor maintains several monitoring locations both within mainland China as well as in Hong Kong.
Track the performance of third party components such as scripts, images and other hosted files. Monitoring 3rd party website elements can help determine whether page load speed issues are affected by content outside of your control. Many 3rd party elements are common resources such as fonts, scripts and images that you rely on the 3rd party to host for high availability and automatic updates. They are also commonly used for links to social media sites or website visitor tracking tools. Many of these third party elements are unpaid resources so it is difficult to hold the providers accountable for downtimes or outages. However, when optimizing the performance of a website, every second counts – which is why many organizations are willing to pay for expedited hosting performance using a variety of methods including CDNs.
Once a website hits a critical mass where simultaneous downloads begin causing slow load speed of individual elements, the next step may be to utilize Content Distribution Networks, commonly known as CDNs. CDNs host your files on multiple nodes around the world, and serve the content up to the end user based upon the closest host or the fastest path. Since CDNs are paid service providers they can be held accountable for the responsiveness of content hosted on the CDN network.
After engaging with a CDN to host your images, scripts and other heavy files, you need to make sure that the CDN is actually providing the speed boosts and scalability they purport to provide. Most CDNs will have a Service Level Agreement (SLA), where they define the minimum thresholds they will provide, including uptime and performance percentages. Monitoring website performance using a 3rd party like Dotcom-Monitor with monitoring nodes around the world gives you an unbiased opinion of how the 3rd party content providers and CDNs are performing from outside your network, the way real users would be experience the content.
Whether your website consists of a simple branding brochure, a complex knowledge base or an exhaustive shopping cart – if the site is offline, it is unable to do its job. Put another way – you will lose 100% of the sales, page views and brand recognition that you could achieve if your website is unavailable.
Furthermore, if the page load time begins to slow and the page performance decreases, your website visitors become frustrated and are much more likely to bounce. According to Kissmetrics, 47% of consumers expect a web page to load in 2 seconds or less. At least 40% of people abandon a shopping cart that takes more than 3 seconds to load, and every second delay can reduce conversion by 7 percent.
First and foremost, you need to know that your website is up. With Dotcom-Monitor it is simple to setup daily, hourly or event minute by minute monitors to let you know if your website or content within the website are down or otherwise unavailable. Once you have setup website monitoring the next step is to observe and visualize the results. Online reports, scheduled emailed reports and automatic email, text message and phone call notifications are available by default.
When configuring website monitoring you can set up alert groups, escalation chains and alerting schedules to make sure the right people are notified of downtime alerts at the proper time. Alert notifications can be sent to one or more contacts as soon as the alert condition is triggered, or you can setup more complex validation rules to rule out false positives and network hiccups. Each period of downtime is also recorded in the reports and can be applied to Service Level Agreement (SLA) reports where you can visualize the percentage of time a website experiences downtime over the SLA period.
Once your custom website monitoring has sent you alert notifications that something is wrong with your website, it is important to quickly and accurately troubleshoot the issue and find a resolution to the problem. Dotcom-Monitor provides a number of easy-to-use troubleshooting tools that can help drill down into errors and identify the root cause of the problem. Due to the complex nature of web applications, Dotcom-Monitor does not provide issue resolution solutions, however you can contact the support team and they will work with you using the built-in troubleshooting tools to identify the source of the errors.
Every time Dotcom-Monitor detects an error or problem with a monitored website, a traceroute is automatically fired off from the location that detected the problem. The results of the traceroute are recorded in several different manors depending on the type of error encountered. A basic traceroute table is recorded as simply as if you ran a command line traceroute from the monitoring node to the target website.
As multiple traceroutes are recorded for a target website, they can be aggregated onto a visual traceroute display that shows how each traceroute traversed from the initial monitoring node across the internet to the target host of the website. This aggregated visual traceroute can be useful diagnosing troublesome nodes across the internet that may be causing packet loss or other timeouts. The latest version of traceroute visualization actually maps the IP addresses of each node in the route onto a global map so that you can see how far the traffic is traveling before reaching the final destination.
Another useful tool that is recorded whenever an error is detected is the actual HTML code sent from your website as well as a screen shot of the rendered page. Part of the website monitoring service includes the ability to see what the website looked like when it was monitored. By default we discard the screen captures for successful monitoring sessions to save storage space, as there is not typically a need to see a successful screen capture. However, each error is recorded and attached to your account for a full week.
Whenever an error is encountered during website monitoring, the specific error code and description of the error are recorded. These details can be found by viewing the online report and expanding the monitoring session to see the individual elements that caused the error.
Waterfall reports can provide some of the most useful diagnostic data by visually displaying the load times and dependencies of each element on a web page. Utilizing a waterfall chart, web page performance can be directly correlated to the slow load times of individual elements. This information empowers you to quickly identify which elements should be optimized to improve page load speed.
An alternative to waterfall chart analysis is to simply click on the element trend tab which displays both the 10% slowest loading elements on the page as well as the 10% fastest elements on the page. Not only will this quickly identify bloated files but it may help you recognize trends. For example, if all content hosted on a 3rd party CDN is listed in the 10% fastest elements, perhaps you should consider moving additional content onto those CDN providers. Conversely, if the 10% slowest report showed many CDN files, testing another CDN provider or even testing serving everything directly from your primary server may be in order.
When you find an element that may be causing problems, click the element trend icon to open up a line chart that displays the performance of only that specific element over time. Ideally this line graph with show a spike at the point where the element started causing problems, so you can identify what was changed at that time. This is an excellent way to analyze historical web performance trends.
Another feature not commonly found on other website monitoring services is video capture, which records a video of the page loading when an error occurs. The video is slowed down and synced up with the waterfall chart so you can walk through the visualization as each element is downloaded and rendered in a real browser. This tool is especially helpful when sharing information about an error or a new bottleneck with team members from another department (such as DevOps to Marketing). Often times the IT team will notice a website suddenly grinding to a halt. Using the video capture paired to waterfall charts they can show Marketing how a very large image file caused a site slowdown and a poor user experience. This makes it much easier to generate quick feedback that results in agreeable changes to optimize file sizes for images, scripts and other large files.
Finally, there are a number of website monitoring reports you can schedule on a daily, weekly or monthly basis which will be delivered to your email as a PDF, CSV or text file. Among those reports is the customizable SLA report which allows you to set your SLA threshold.
Website monitoring is essential to running a successful website. Monitoring lets you know if there are problems with your website, so and can quickly diagnose and repair performance issues you would have otherwise missed. Website monitoring can be as simple as a ping check, or as complex as a scripted walk-through of every single page on the website. The complexity of your website monitoring is predicated upon the importance of each page to the health and value of your website. Some organizations may only need to check the performance and availability of their landing pages a few times a day to ensure the pages are up, while other high performance SaaS models might need minute to minute website performance metrics in order to provide the best possible customer experience.
Without proper server health monitoring, your website may be slow, may buckle under heavy load, and it may crash. Services like Dotcom-Monitor continually check your web servers at a specified interval to ensure availability, proper response time, and even text-based content verification. Alerting and reporting enables you to quickly diagnose and resolve the root cause of any website or server issues.
It is essential to monitor your email systems to be alerted the moment a problem is detected with your email server. Dotcom-Monitor provides a number of email monitoring tools from SMTP server monitoring, to Microsoft ActiveSync monitoring and even round-trip email monitoring, which will both send and receive an email from your server to verify the health of the system. You can also monitor Outlook Web Access (OWA) with the website monitoring platform.
Actively manage VoIP servers with Dotcom-Monitor VoIP-SIP monitoring. Get alerts on availability or performance issues with your VoIP communications system. Dotcom-Monitor's VoIP-SIP monitoring replicates calls to an end-user's SIP device and analyzes the call responses to determine the connectivity of VoIP services. When a problem is detected an instant alert is sent, detailing the issue. This empowers your telecommunications and IT teams with the right info to quickly resolve issues and reduce mean time to repair (MTTR).
These days, the lines are blurring between web sites and web applications. Many people use SaaS applications and don't even think about them as a website. Conversely, many people have never considered their website to be an application. With new Rich Internet Applications (RIAs) being developed every day on top of solid interactive frameworks like AJAX, jQuery, AngularJS, Node.js, HTML5, Flash, Java, Flex, Silverlight and the likes, traditional desktop based applications are moving into the web app space. It is crucial to have a web application monitoring system in place that can run multi-step scripted scenarios in real browsers to test real user scenarios.
Dotcom-Monitor has a comprehensive suite of services, from monitoring simple HTTP GET and POST requests using headless browsers, to fully scripted real browser interactions with video recording. The most basic of these website monitoring systems is the ServerView HTTP/S monitor. You simply specify a URL to hit, identify whether basic authentication is required – SSL or TLS – and determine how much of the web page to download. The basic ServerView HTTP task only sends the initial GET request to download the HTML. More advanced HTTP tasks can also request to download all subsequent files specified in the initial HTML download.
BrowserView performs full page monitoring by downloading and rendering single pages in a real web browser to detect page performance. Select from over 40 different browsers or mobile devices to monitor the performance of a single webpage. This website speed monitoring continually checks web page load times in Internet Explorer, Chrome and Firefox from nearly two dozen global locations, ensuring your load times are within your set threshold. This ensures a consistent and positive experience for your users.
To gather performance metrics for complex web applications you may need to perform several scripted actions in sequence, such as logging in to the application, performing a task such as a search, or add an item to a shopping cart – followed by a checkout or log off. UserView uses the EveryStep script recording tool to record such scripts and perform website monitoring similar to the full page download and rendering of BrowserView. This adds the advantage of stringing multiple pages together and recording user interactions on each page.
UserView is capable of recording button clicks, data entry into forms, and even user actions on interactive content such as AJAX, jQuery, AngularJS, Node.js, HTML5, Flash, Java, Flex, Silverlight and other RIA components. Validate the content, functionality, and visual layout of web pages, websites, and web applications – no matter the language used to code them.
Schedule load tests to proactively perform capacity planning based upon expected user load during peak days and times of the year. Running such proactive performance plans can help establish baseline performance expectations, particularly when you are about to release a new product or campaign. Stress tests are also an important part of a healthy website monitoring toolbox because you see the performance of your website and infrastructure under extreme conditions. By stressing your infrastructure to its limits you can see if it is capable of handling the load gracefully or if things begin to falter and crash. Load testing can be run via headless browsers – similar to PhantomJS and Selenium, but with our own proprietary scripting browser – or via real browser testing, which renders and interacts with the website for greater user experience testing.
Any third party content used on a website poses a potential risk because the source of the content is not under your control. Therefore, it is in your best interest to monitor such elements to ensure that you know if the element ever breaks or if the load speed of the element slows down and becomes a significant bottleneck in the speed of your website.
CDNs are often great for speeding up and optimizing the performance of a website, but you need to monitor the performance of the CDN content not only from your own network, but from different locations around the world. By doing this you are effectively testing the propagation of content on the CDN to the nodes closest to your website visitors and ensuring that each geographic region receives the same user experience regardless of how far away they are from your web servers.
As the use of online web based applications continues to grow, SaaS providers need a way to monitor their web application performance externally. Dotcom-Monitor is an unbiased 3rd party that provides performance metrics for complex use cases of web applications. Once acceptable baseline performance metrics have been recorded you can establish a limit for acceptable web app response time. Once that threshold is breached, the website monitoring system will send alert notifications when the web app is not performing within these guidelines.
** Try the comprehensive Dotcom-Monitor suite of website monitoring tools today with a Free 30-Day Trial. **