|
We are often asked about how many URLs should be monitored. The flip, but
correct, answer is that it depends on what you are trying to do.
As general guidance, we suggest that institutions track at least
those URLs served by different Internet servers or supported by different back
end systems. For example, an institution might have its main site hosted on a
hosting service, its Internet banking site hosted by another, its check reorder
feature outsourced to a partner, and its credit card products outsourced to
another partner. We would suggest that this institution monitor a URL on each of
these subsystems.
Notwithstanding this guidance, there are many good reasons for
monitoring more. If you are concerned, give us a call and we will help
work out what's right for you.
You can change the URLs you monitor whenever you wish. As a practical matter
given the reporting structure, it generally only makes sense to change URLs at
the end of a reporting period. There are several good reasons for this, not the
least of which is that the reports only report on the URLs active at the time
the report is generated.
We generally discourage changing the URLs frequently to test more of the
site. While you can do so, as a practical matter it is usually much more useful
and ultimately less expensive for you to simply purchase additional URLs.
Each service level comes with a starting number of URLs. The number is chosen
to be consistent with the reports and analysis provided. For many users these
URLs are all they need. However, if you need additional URLs you can add them to
any service level.
As a general rule if you need more than one additional URL, it is
advantageous to upgrade to the next service level.
If you need to monitor many URLs, look into
100% Sampling.
One exciting option we offer as part of Monitor3
and Monitor4 is 100% Sampling.
100% Sampling lets you monitor your entire site to insure that
your customer's Internet experience is positive wherever they go on your site.
This option recognizes that there are some URLs you need to continuously
monitor and others that you can check much less frequently, that you can sample.
With 100% Sampling we will monitor as much of your Web site as
you wish, including all of it.
It works like this.
- You provide us with one or more starting URLs.
- We build a list of all the URLs on your site (we do not cross domains)
and send it to you.
- You select the URLs you want to test less frequently and send the list
to us.
- We add these URLs to the sampling URL list.
URLs on this list are tested on a time available basis. That is, we first
test all the URLs on the continuous monitoring list, and then if there is any
time available before we have to start working on the continuous monitoring list
again, we test URLs from the sampling URL list.
How often these URLs get tested depends on many factors, most of which we
have little control over, so it is difficult to predict how often these URLs
will be tested. As a practical matter, however, in our experience they are all
tested several times a day.
We do not send email alerts for URLs on the 100% Sampling
list. While it is technically possible to do so, we are reluctant on the
argument that sending email alerts would encourage a level of confidence in
availability and performance not supported by the sampling technique.
Additionally, 100% Sampling URLs are only tested using the
single test protocol.
100% Sampling URL results are contained in a separate report
we send to you quarterly. There are several reasons for this. The most
significant is that since the URLs are tested much less frequently (a few times
per day as compared to 48 times per day for continuous monitor URLs) it takes
longer to accumulate enough data to have meaningful results. Additionally, the
report contains somewhat different analysis consistent with the irregular
sampling interval.
We are occasionally asked "Why do 100% Sampling?" This is
often coupled with "Why not just use a link checker?" There are several reasons
that 100% Sampling makes sense. The first has to do with what a
link checker does versus what we do. A link checker works by parsing the content
of a URL (HTML), finding all the links, and then opening and reading their
content. If that content in turn has links, those are followed until all links
have been found and found to be readable. But readable and browse-able are,
unfortunately, often not the same thing. Content that can be read may or may not
actually be something that a browser can render. Additionally, link checkers can
not measure real response times. Conversely, we use a browser to access the URL
and measure the response time. If this test succeeds there is no question that
the page is usable.
There is a second, and in many ways more important reason, to use a browser
to test the page. While you probably have good control over your own domain's
pages, you probably have very little control over your business partners' pages.
If you can link to a page on your site, then chances are it is browse-able, and
if it is not you can generally do something about it since you control the site.
Not so with a partner's pages where you have much less or no control over the
pages. The only real way to determine if a partner's page is good is to actually
browse it.
We are occasionally asked if we will include pages from other domains in the
100% Sampling list. The answer is it depends. Many such
situations are entirely benign. Others need to be reviewed. As a general rule we
will check a few pages on the sites of business partners your site links to.
However, if you want a more extensive analysis of your business partner's site's
performance, we need to review the situation. If you have a concern in this area
let us know and we will work with you to meet your needs.
100% Sampling is available only to Monitor3
and Monitor4 customers as this service requires
significant setup and management. The cost of 100% Sampling
varies with the number of URLs on the list. Typically the cost is in the range
of $1200 to $3600 per year for a small to medium site (100 URLs to 300 URLs). If
you are interested, let us know and we will quote to your specific needs.
The easiest URLs to monitor are those with a constant HTTP address. Rarely,
we encounter a URL that for some reason we have difficulty measuring. Typically,
this is due to page scripting or other sophisticated features. As a rule, we are
able to resolve these. We are continuing to work on our monitoring facility with
a view to being able to monitor any page. At the present time we are monitoring
URLs on 203 of the top 206 banking Web sites successfully. You can find a list
of many of the sites we have monitored
here. If you have any concerns in this area, provide us
with the URLs and we'll verify that we can monitor them.
Transaction URL Test is available to Monitor3
and Monitor4 customers. With the Transaction
URL Test option, Monitor will log into a site and
perform a predefined sequence of screen interactions. As these interactions are
performed, Monitor measures the site's transaction response
time. Transaction sequences are customized to meet the needs of each customer.
The cost of Transaction URL Test varies with the number of
URLs in the sequence, the sequence complexity, the sequence stability and other
factors. The cost is typically in the range of $200 to $800 per transaction
sequence per month on a 12 month agreement. | 
URLs
