Website availability

Website availability (also called website uptime) refers to the ability of the users to access and use a website or web service. A website’s availability is typically communicated as a percentage for a given span of time.

What is meant by “website availability?”

An available website is a website that is accessible and usable as expected by the user. When speaking about availability, we often refer to the ratio of the available time to the total time. Typically, a site’s availability is described as a percentage such as 99.9% availability.

For a site to have 100% uptime is an elusive, if not impossible goal over the long term. Most providers often have an availability goal they try to reach and maintain. Let’s take a look at how to calculate the uptime percentage.

Calculating website availability

Availability is a simple percentage based on the uptime divided by the total time span. Depending on the degree of accuracy you need, you may use hours, minutes, seconds, or milliseconds. For our example let’s use hours.

A site with one hour of downtime during an entire year would have a 99.99% uptime.

((8760 hours – 1 hour)/8760 hours) * 100 = 99.99%

99.99% uptime sounds impressive with only one hour of unplanned downtime per year, and for most sites, this is an excellent goal. Other sites like Google work to exceed a 99.999% availability (5.26 minutes of downtime per year).

Does available only mean accessible?

For many brands promising high availability, they only refer to the time that a user can enter a URL and not get a page error such as:

• 404 page not found
• 500 internal server error: A generic code for an unknown error occurred.
• 503 service not available: Refers to a temporary problem.
• 504 Gateway timeout: a server didn’t respond in a timely manner.

Of course, you may encounter many other errors as the result of an outage. The simple fact is that the user can’t access the site or service for some reason beyond their control. Does availability only mean the ability to access a page, or does it include the ability of the user to complete the task? That is a good question, and the answer varies based on who you ask.

Is a malfunctioning website still considered available?

If you ask the user if a malfunctioning site is available the answer is probably, “No.” If you ask a hosting provider the answer is probably, “Yes.” So, which is it? Based on the definition above, if the user can’t get the information they need or can’t complete a task, you should consider the site down. To capture this type of outage, use Transaction Monitoring.

Is a malfunctioning website worse than a full outage?

Although it’s true that an outage blocks 100% of users, studies show that 91% of those users will return to the site. For a site with issues but still accessible, the permanent abandonment rate is 28%. The study based its findings on short term outages, so the longer the outage lasts or more frequently outages occur, the more impact outages have on reputation and revenue.

Does website performance affect availability?

It can if the poor performance prevents the user from completing their task. So, based on our basic definition of “availability,” sure performance is also an availability factor. The human brain needs fast responses to maintain focus when it comes to human/computer interaction. Slow responses break concentration and negatively impact user perception of the brand (Everts, Tammy. Time is Money). Plus, slow load times on mobile also results in lowered search rankings on Google.

Poor performance is contributable to poor design and setup with the number one problem being large images. Besides image optimization, consider asynchronously loading render blocking content, using compression, caching, browser hints, keeping third-party content in check, and using CDNs to reduce latency to improve your website’s performance.

Is planned maintenance considered an outage?

Typically, periods of planned maintenance are not considered downtime. Usually, during the maintenance period, developers redirect the user to a page that apologizes and explains that the site is under maintenance and to please try again a little later rather than returning an error code.

With many cloud-based services, the content is distributed over multiple servers, and pulling one server for maintenance just redirects requests to other available servers thus, avoiding downtime. Redundancy is your biggest defense against outages, but it comes with a high price tag that is well worth it for the big players like Amazon who generate hundreds of thousands of dollars per minute, for the rest of us and comprehensive monitoring strategy is your best option for maintaining high availability.

Tracking website Availability

To track availability, a brand needs to check to make sure that users can connect error free to their website using regular, frequent test intervals. Some brands check manually periodically through the day, but short outages may go unnoticed by periodic testing. Automated testing can test for uptime as often as every minute, so even those short outages that affect their users are captured in uptime reports.

How does website availability monitoring work?

Availability monitoring is a proactive option that verifies the response from your site using servers external to your site. Using external servers, preferably servers located near your actual users, you capture uptime and latency issues based on your users’ locations. Here’s how it works.

1. A remote test server sends a request to your page for content.
2. The test server receives a response code of “200 ok,” and the site is determined to be up or available.
3. Another test server in a different location sends a request 60 seconds after the previous test completed.
4. The test server receives an error code.
5. Another test server immediately sends a request.
6. The second test server receives an error.
7. The site is considered down at the time of the first error.
8. Testing resumes at one-minute intervals.
9. A later test gets a successful response.
10. The site is considered up, and the monitoring service records the time between the first error and the successful test as downtime.

Alerts go out after the second confirmed error to let you know your site has an availability issue, and an all-clear alert goes out when you’ve resolved the problem. Besides a positive response (200 ok) the availability checks can also verify:

• Response time
• Response size
• Specific HTML result code
• Specified content

Discover localized availability problems

Did you know that outages may affect some users and not others? Due to local restrictions, poor local Internet architecture, network issues such as DNS problems, browser type, and device type some users may not have access to your site. Using a monitoring service with a large checkpoint network, you can select test locations more representative of your user base allowing your monitoring to capture localized downtime and latency issues.

Conclusion

Availability is more than just uptime, but without uptime, the other problems affecting availability don’t matter much. Availability Monitoring is fundamental to a sound digital experience monitoring strategy.