Amazon AWS S3 outage

March 1st, 2017
Posted in Monitoring, Industry News

Amazon AWS S3 outage


Yesterday Amazon's S3 web-based storage service experienced a major outage, which left a lot of websites and applications not working properly or not working at all. Not all Amazon customers were affected, but a lot of sites relying on storage in the US EAST region experienced timeouts or loss of functionality. Among those sites are Giphy, Medium, Slack, Quora and DownDetector.

Below is the offical timeline of the outage from the Amazon’s AWS service health dashboard. As you can see even the Amazon's own health dashboard was malfunctiong due to this problem

[RESOLVED] Increased Error Rates for Amazon S3
Update at 2:08 PM PST: As of 1:49 PM PST, we are fully recovered for operations for adding new objects in S3, which was our last operation showing a high error rate. The Amazon S3 service is operating normally.
Update at 1:12 PM PST: S3 object retrieval, listing and deletion are fully recovered now. We are still working to recover normal operations for adding new objects to S3.
Update at 12:52 PM PST: We are seeing recovery for S3 object retrievals, listing and deletions. We continue to work on recovery for adding new objects to S3 and expect to start seeing improved error rates within the hour.
Update at 11:35 AM PST: We have now repaired the ability to update the service health dashboard. The service updates are below. We continue to experience high error rates with S3 in US-EAST-1, which is impacting various AWS services. We are working hard at repairing S3, believe we understand root cause, and are working on implementing what we believe will remediate the issue.

Some of our clients experienced issues due to this problem, but we were able to quickly identify the issue. Switching to other working platforms prevented huge losses of revenue and decreased their downtime.

The Free Website Test Tools by WebSitePulse

May 31st, 2016
Posted in Monitoring

The Free Website Test tools by WebSitePulse allow you to test a server, website or network as a one-time thing, without having to create a monitoring target. We have tried to keep them as simple and easy to use as possible. Here is a quick guide to how each test can help you.

The test tool section is divided into 4 sub-sections: Website testing, DNS testing, Service and Network testing.

In this article we will concentrate on Website testing.

First of all, keep in mind that if you are a new user and do not have a registration with us, only three monitoring locations will be available. After you sign up for our services and create a monitoring target, you will be able to see the list of all our monitoring locations.

The Website testing page has five tests available:

Website Test

How does it work? Firstly, we verify the server status. After an OK status is returned, we download the full HTML content and check the response time of the tested website. The results also provide you with the DNS lookup time, connection time, download the complete HTML of the tested website and first and last byte download time. Here is a quick preview:

Click to Enlarge

Web Page Test

With the Web page test, we verify the URL availability, then download the complete HTML content and all of the images and other internal web page objects. After that we measure the download speed of each of the components. The test results display the DNS time, connection time, first and last byte times for each of the web page objects. Keep in mind that this test is limited to 50 page objects. You can also choose the type of chart displayed - You can review it in Flash, Java or PNG image. Or, if you would like, you can exclude the graphic from the result page and show only the data and the internal references. Tip: If you want to email the test results, you should select the PNG chart. Here is how it looks:


Click to Enlarge

HTTP Headers Test

​What the ​​​h​eaders test does, is​ requests the entered URL and ​retrieves the HTTP response headers. Afterwards​​,​ it verifies the HTTP status codes and displays the received response headers. You can find out more about the different HTTP status codes here

Click to Enlarge

Link Check*

The Link check test first verifies the URL availability and downloads the complete HTML content. After that, it checks the availability of the internal components and of all of the links which are found present in the page you are testing. After it is finalized, the test displays DNS time, connection time, the first and last byte times for all components of the web page and the DNS, as well as connect and response times for the links from the tested URL. This test is also limited to 50 page objects.

*Note, that this test works only if you are a WebSitePulse user, and you have an active target in your account.

Click to Enlarge

Test  Behind the Great Firewall of China

Many of our customers have stumbled upon this problem - their website cannot be accessed from China. Now, why is that?  Basically, 15 years ago, the Chinese government started the "Golden Shield project", a.ka. The Great Firewall of China. Its main purpose was, and still is, to filter literally all the traffic coming in and out of China. You can read more about the Great Firewall here: and here: With our test tool you can check any website in real-time to see if it is accessible from China. We provide testing your website from all our China monitoring locations: Shanghai, Beijing, Guangzhou and Hong Kong.

Handling Downtime and Outages

May 30th, 2016
Posted in Monitoring

We all have experienced outages and downtime and we all know that this has quite a negative impact on our business and reputation. Unfortunately, downtimes and outages cannot be completely avoided, but can be quickly detected, thus minimizing the negative effects. All we need is a very good tool, which will help us identify even the shortest server overloads or network glitches.

First of all you have to be prepared to handle the situation. Knowing the configuration specifics of your website / server / third party hosting panel is crucial if the problem is on your side and can be fixed by you. There are times when you will not have control over the situation as the issue is with a major upstream provider, but being notified immediately about it gives you the chance to contact the provider and let them know that there is a problem. Using our services gives you another option – our Takeover feature. It allows you to leave detailed instructions that our support team will follow in case of a detected failure. You might want us to contact your hosting provider and report the problem, give us credentials to SSH to your server and manually reboot it (if possible), etc.

Click to Enlarge


There are times when you need to do some updates or perform other mandatory changes to your website / server, which require you to manually and intentionally put it them in a maintenance mode. This is not considered a downtime and you will most probably not want it recorded as such. We have thought of that and we have the Scheduled Downtime feature for you. You can specify certain periods during which we will not perform any checks, thus keeping consistent and high uptime figures.

Click to Enlarge

It is also very important to have your alerts configured properly. Make sure that you have selected all alert types (Connection related alerts, Content related alerts, Timeout warnings and Recovery messages) to be notified for all potential problems. The Failures before sending alert value is also recommended to be set to 1, so that you can be alerted on the first detected failure. You do not have to worry about false-positives as we provide a second backup location for each of our primary ones. If the primary location detects a problem, the backup one immediately performs a second check and an alert is only sent if there is a confirmation.

We support a wide variety of contact types, which gives you the flexibility to choose the best option for you. The contact types that our system supports are the following:

  • SMS contacts
  • E-mail contacts
  • Voice call contacts 
  • iOS and Android PUSH notification contacts
  • Instant messenger contacts
  • URL contacts
  • SNMP contacts

Click to Enlarge

Keep in mind that outages and downtime are inevitable and you will deal with such a problem eventually, but being prepared can make all the difference. Once the problem is fixed, it is very important to understand what caused the issue and how to prevent it from happening again in the future. We can help by providing very detailed reports, as well as PING and Traceroute information (if enabled) for the specific checks. Content issues can also be captured with some of our more advanced levels of monitoring – Performance, Full-Page and In-Browser ones.

Monitoring Outlook Web App (Case Study)

October 13th, 2015
Posted in Monitoring

What is Outlook Web App (OWA)?

Outlook Web App (OWA), first released as "Exchange Web Connect" and then renamed to "Outlook Web Access", is a webmail interface for Microsoft Exchange Server.

OWA was created in 1995 by MS Program Manager Thom McCann from the Exchange Server team. The first customer version was provided as part of the Exchange Server 5.0 release two years later.

Microsoft released the OWA customer version as a component of Exchange Server to allow users to use their email accounts via a web browser without installing Microsoft Outlook. OWA provides features for managing contacts, documents, calendars, tasks and other mail-related content.

Outlook Web App has had two user interfaces available since its release back in 2000 - one with all features included ("Premium") and one with а limited functionality ("Light" a.k.a. "Lite"). Prior to the 2010 release, OWA required full functionality only from the Internet Explorer, but today it supports most of the commonly used browsers.

Advantages of OWA

The Outlook Web App:

  • Provides quick and easy way to do your email, calendar and contacts from any device (PC, mobile phone, tablet, etc.)
  • Allows access to the user account from any computer outside the local/office network
  • Runs on most Internet-enabled devices
  • Offers light version for users with bad Internet connection
  • Supports S/MIME
  • Has anti-phishing and anti-spam technology
  • Provides instant messaging

Many of the OWA users are companies with lots of personnel who use OWA as a fast and easy way to access their email accounts from virtually everywhere and from any device.

How Can WebSitePulse Monitor OWA

WebSitePulse can offer the following monitoring solutions for Exchange Serverand particularly Outlook Web App:

  1. Email Round-trip Monitoring
  • It’s easy to setup
  • It checks the incoming and outgoing email flow
  • It does not check the OWA web interface
  1. Front Page OWA Monitoring - monitor the OWA front/login page to see if it is working properly with ourPerformance and Full-page monitoring levels.
  • It’s relatively cheap
  • It supports lower monitoring frequencies (1/3/5 minutes)
  • It does not verify the proper functionality of the customer's email system
  1. OWA Login - monitor the OWA to see if it is processing the logins properly andshowing the mailbox home page with our Transaction monitoring tool
  • It verifies if the OWA is actually accessible to the users
  • It does not check if the OWA is able to send or receive emails
  1. Full OWA Functionality Test - with our Custom Transaction Monitoring, we are able to test the full OWAfunctionality. This includes the proper login processing, the ability to sendand receive emails, the functionality of the other major OWA features (search, Calendar, Contacts, Tasks, etc.).

Case Study

Here is an example of a monitoring solution we have developed for one of ourcustomers. We use our Custom Transaction Monitoring to test the client'sOWA capabilities of sending and receiving mails.

The test process starts by sending a unique email to a customer's test emailaccount. Then, our monitoring system logins into the OWA and searches for thatemail. If the email is present, it is being deleted and then a new uniqueemail is being sent through the web interface to a mailbox outside thecustomer's network (in this case - a WebSitePulse mailbox). On the finalstep, our system checks if the second email has arrived.

This custom monitoring solution verifies if the client's email system (inbound/outbound email flow) and OWA are working properly.

Apart from the functionality test, we also measure the OWA interfaceperformance.

The customer will be notified in any of the following cases:

  • The OWA is not loading at all or is performing slowly
  • Our system cannot login with the provided user credentials
  • The incoming test emails are missing
  • Our system cannot send emails through the OWA web interface
  • The outgoing test emails have not been delivered
  • Issues with secured OWA zone appear

This monitoring solution has been developed with a 9-step custom scripttransaction. The deployment time for such monitoring is usually 24 hours andthere are no setup fees.

I Didn’t Get My WebSitePulse Alert! Why?

October 5th, 2015
Posted in Monitoring

One of the main reasons you may start looking for monitoring tools is the need to get notified when your website is not performing properly.

Of course, you need to be aware of the problem before your customers do so you could take actions to avoid customer and revenue loss. After all, if your website has been down for a while without your knowledge, you are likely to experience serious business downfalls. To avoid critical situations like that, you should simply ensure you have set the alerting of your monitoring targets properly.

How Does WSP Alerting Work?

There are three main sections in the target configuration responsible for the proper alerting.

Target Status

Click to Enlarge

A single click on the target label will lead you to the target status – a place where you could check your current target status and a focal point to all of the target’s features. Once you reach the target status, you could check if the alerts of the target are active or suspended. It’s that little speaker icon on the right-hand side on the top of the screen. If the target alerts for the respective target are active, you could proceed to the target settings.

Alerts Configuration

Click to Enlarge

This is where you set the existing notification contact to receive alerts for the target. WebSitePulse splits the errors into three main categories, which is why the alerts are combined in three groups – connection-related alerts, content-related alerts and timeout warnings. Depending on your needs, you could use one, two or all three options at the same time. Note that if you leave the timeout checkbox for a specific notification contact unchecked, the same notification contact won’t be alerted with a timeout warning if an issue occurs.

There are several more columns in the alerts configuration table. One of them is failures before sending. The default value here is 1, meaning you will be alerted on the first detected failure. You could change that value as per your needs, but keep in mind that if you set it to 2, once we detect a failure, we will wait for the next check and will send alert to the notification contact only if the problem is detected again on the next (second consecutive) check. This feature is commonly used as a false-positive prevention system for targets with short monitoring intervals (1,2,3 minutes) and it’s not recommended for targets with monitoring intervals higher than 5 minutes.

Monitoring locations

Click to Enlarge

Here you select the location from which your target will be monitored. If the alert settings’ option is set to “don’t send”, we won’t send you an alert even if the monitoring location detects an issue.

While you are in the monitoring location section, you could also check the error confirmation feature. When this feature is activated, we run a secondary check from a different location to confirm the error before sending you an alert (when a problem is detected). There are cases when the email report shows you a failure for which you haven’t receive an alert. An unconfirmed error could be one of the reasons.

Now, take a look at the location-correlated notification level feature. Again, this is our false-positive prevention system, working only when your target is monitored from more than one monitoring location. If your target is monitored from three monitoring locations, this feature could be set to alert you only if at least half of the monitoring locations (in that case - at least 2 of them) detect the same error. In certain cases, you could increase the correlation level between the monitoring locations and set them to alert you only if all of them detect the same error.

Other reasons which could prevent our system of sending you an alert are:

  • An active dependency rule for the target
  • An active contact availability rule for the notification contact

These are exceptional cases for which we assume you are aware of since you have manually set them up.

If all features and settings described above are set properly and you still have not received your alert, contact our Support team.