Hi,
Boomerang seems a nice option, but I am working in a cloud environment and
Boomerang requires javascript to be embedded in every page that needs to be
monitored. The applications are run on the lighttpd servers by the users and
it might seem intrusive to embed javascript into their application without
their consent in the cloud environment.
Piwik is another tool which also provides client perceived response time
measurements but with the usage of javascripts similar to Boomerang.
Does anybody know any other tool for client perceived response time
measurement without the use of Javascripts?
-Archana
On Sun, Mar 13, 2011 at 8:24 PM, Vladimir Vuksan <vli...@veus.hr> wrote:
> There are multiple levels of looking at performance. Tool I wrote analyzes
> server side performance since that was the main goal of project however you
> could adapt it to analyze any other type of response times ie. there is a
> parser that takes URLs and how long it took the web server to serve that
> request. User perceived performance is much harder to gauge since it depends
> on a number of factors not under your control such as CPU speed, rendering
> engine, spyware etc. slowing things down etc. You should obviously get as
> many metrics and evaluate how you can speed things up but server side
> performance ie. time it takes to generate the HTML is one you have most
> control over.
>
> As far as client perceived response time there is also Boomerang
>
> https://github.com/yahoo/boomerang
>
>
> Vladimir
>
> On Sat, 12 Mar 2011, Archana N wrote:
>
> Thank you so much. This looks really interesting. I am working on a
>> project to monitor the performance of lighttpd servers in a
>> cloud environment(for resource provisioning purposes) . Also, in the blog,
>> I was not clear by the term "response time". Is it the
>> response time as perceived by the client or is it the request processing
>> time of the server. I had seen many tools such as Piwik
>> which would measure the client percieved response time of the web pages,
>> but they have the disadvantage of having java-script to
>> be enabled at the client's for this purpose. Is the tool you had developed
>> similar to Piwik or is it measuring the server side
>> processing time.
>>
>> -Archana
>>
>> On Sat, Mar 12, 2011 at 9:46 PM, Vladimir Vuksan <vli...@veus.hr> wrote:
>> You could use Ganglia but depending on number of unique URLs this may
>> not be such a great idea. I did something
>> similar at a previous job to evaluate page response times (aggregated
>> on hourly basis). You can find it here
>>
>> https://github.com/vvuksan/pagetime-analyzer
>>
>> I blogged about it here
>>
>>
>> http://vuksan.com/blog/2010/07/15/analyzing-your-web-page-response-times
>>
>> Vladimir
>>
>>
>> On Sat, 12 Mar 2011, Archana N wrote:
>>
>> Hi,
>> I was also thinking of having metrics for counting the number of
>> access of a particular string, but the problem
>> is I am working
>> with an application which has many directories (similar to wordpress)
>> and there are a lot of directories which
>> get different
>> amount of hits by the users and I would like to track which ones are
>> mostly hit. I will also try to think about
>> it. I was not
>> sure if Ganglia provided the aggregation, but now I got the answer to
>> that :) .Do let me know if you also come
>> up with ideas.
>>
>> On Fri, Mar 11, 2011 at 6:34 PM, Rick Cobb <rick_c...@ieee.org>
>> wrote:
>> Ganglia doesn't have a model for aggregating string-valued
>> metrics.
>>
>> On the other hand, you can get a long way by having your
>> metrics-gathering modules post a count *per string* (e.g.,
>> "www.yahoo.com:hits", 15); you'll have a ton, though, so you
>> may want
>> to use some sort of naming prefix to help organize them.
>>
>> If you have ideas about how you'd like to see them aggregated, I
>> for
>> one would love to hear them; it's a fun problem to try to solve.
>>
>> -- ReC
>>
>> On Fri, Mar 11, 2011 at 9:13 AM, Archana N <dreamgirl...@gmail.com>
>> wrote:
>> > Hi,
>> > I am using Ganglia for monitoring lighttpd server statistics in a
>> cloud
>> > environment. In my case I would have string metrics such as the
>> pages which
>> > are frequently accessed etc. I understand that if there are many
>> clusters,
>> > then Ganglia aggregates the information at the grid level. This is
>> possible
>> > for the numerical metrics. However, I would like to know how this
>> would work
>> > for the string metrics.
>> > -Archana
>> >
>> >
>> ------------------------------------------------------------------------------
>> > Colocation vs. Managed Hosting
>> > A question and answer guide to determining the best fit
>> > for your organization - today and in the future.
>> > http://p.sf.net/sfu/internap-sfd2d
>> > _______________________________________________
>> > Ganglia-developers mailing list
>> > Ganglia-developers@lists.sourceforge.net
>> > https://lists.sourceforge.net/lists/listinfo/ganglia-developers
>> >
>> >
>>
>>
>>
>>
>>
>>
>>
------------------------------------------------------------------------------
Colocation vs. Managed Hosting
A question and answer guide to determining the best fit
for your organization - today and in the future.
http://p.sf.net/sfu/internap-sfd2d
_______________________________________________
Ganglia-developers mailing list
Ganglia-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-developers
