topic Re: URL report top 100 with browsing time in General Topics

URL report top 100 with browsing time

TimmyLamar — Wed, 09 May 2018 10:43:21 GMT

Hi All,

i was trying to get as close as possible to bluecoat report that able to provide

1.URL visited

2.URL category

3.Browse time

4.User

So the closest i can get is URL report with counts, but not browse time,

secondly i can get Traffic logs with elapsed time, without URL info.

is there anyway i can mix them with api function?

thanks in advance

Re: URL report top 100 with browsing time

Remo — Wed, 09 May 2018 19:15:49 GMT

If you are willing to write a script, then yes this is absolutely possible with the API. (Very simplyfied) pull the URL Logs and then get the elapsed time from the traffic log - the "primary key" here is the session ID. With the session ID you can mix them together.

And now my personal opinion to reports with browse time: they are crap. There is almost no way to really distinguish between the actual time the user actively spent with web-browsing and background browser sessions. Connections to modern (ajax based, heavily javascript dependant, dynamic) websites stay open even when the user isn't actively viewing the website, while for example a user could spend hours to read a big wikipedia article but in the report you only see a few seconds as the session closes when the content is downloaded. In this example the other employee who opens facebook for like 2 minutes in the morning, but does not close the tab will probably show up in the report that he is using facebook 9/5.

As I have never actively used a bluecoat product, this may not apply to this report. As I wrote it is almost impossible to do such a report. With heavy data analytics or if bluecoat injects a javascript into every web browsing session it gets feasable to have browsing times that are way more realistic than a simple session time report.

Re: URL report top 100 with browsing time

BPry — Wed, 09 May 2018 19:40:58 GMT

@TimmyLamar,

We see requests like this a lot and everything that @Remo said is true, he even gave you way more information on how you would do so than what is usually given. Generally the consessus that everyone comes to is that there is no way to provide a report like this with any amount of viable accuracy.

If the request is because someone wants to view the information and fully knows that it may not be accurate and that they can't make any determinations on it, then fine. If this is a request because they want to get HR envolved or sit the employee down because they forgot to close a browser tab, then it's better to simply not provide the report at all.