You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@jmeter.apache.org by "J.E.C.Brisland" <j....@open.ac.uk> on 2010/11/05 12:57:21 UTC
Best way to build a test...
Hi All,
Basically what I want to do is reproduce the average load our server has on a daily basis onto our test server.
We have a number of different page types on our server, and each of these types gets hit a varying amount.
Take for example blog views, posts and comments.
Lets say these are our numbers
Per second
0.5 users log in
3.4 blog views
0.2 blog posts
0.002 blog comments
What would be the best way to reproduce this load using jMeter?
I have tests running, but I'm struggling to reproduce the loads I want.
--
The Open University is incorporated by Royal Charter (RC 000391), an exempt charity in England & Wales and a charity registered in Scotland (SC 038302).
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: Best way to build a test...
Posted by Felix Frank <ff...@mpexnet.de>.
On 11/05/2010 01:07 PM, J.E.C.Brisland wrote:
> Note: I also have a whole bunch of other things I want to add, for example
>
> Per second
> 4.8 forum views
> 0.2 forum posts
> 0.9 general page views
> Etc..
>
> The concurrency of these requests is not known, so I just want to randomise it a little, as long as I can get a general load per second it doesn't really matter (to me anyways) about overlap. If it's random it will be good enough.
>
> -----Original Message-----
> From: J.E.C.Brisland [mailto:j.e.c.brisland@open.ac.uk]
> Sent: 05 November 2010 11:57
> To: 'JMeter Users List'
> Subject: Best way to build a test...
>
> Hi All,
>
> Basically what I want to do is reproduce the average load our server has on a daily basis onto our test server.
>
> We have a number of different page types on our server, and each of these types gets hit a varying amount.
>
> Take for example blog views, posts and comments.
>
> Lets say these are our numbers
>
> Per second
> 0.5 users log in
> 3.4 blog views
> 0.2 blog posts
> 0.002 blog comments
>
> What would be the best way to reproduce this load using jMeter?
>
> I have tests running, but I'm struggling to reproduce the loads I want.
>
For each scenario, create one Thread Group that does all involved
requests. To each Thread Group, add a Constant Throughput timer that
ensures that the desired number of transactions is maintained.
Regards,
Felix
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: Best way to build a test...
Posted by Evandro Grezeli <gr...@gmail.com>.
Buddy,
If i didn't get wrong, you want to simulate the think time of your step's.
I dunno if you are the user or you just have those information from access
log's, but why you don't record your script using Gausian timer (With the
${T} variable) to exactly record what do you want?
Or perhaps yous could record your scenario without using the gausian timer
and just set those time in each step.
My 2 cents.
[]s
On Fri, Nov 5, 2010 at 10:07 AM, J.E.C.Brisland
<j....@open.ac.uk>wrote:
> Note: I also have a whole bunch of other things I want to add, for example
>
> Per second
> 4.8 forum views
> 0.2 forum posts
> 0.9 general page views
> Etc..
>
> The concurrency of these requests is not known, so I just want to randomise
> it a little, as long as I can get a general load per second it doesn't
> really matter (to me anyways) about overlap. If it's random it will be good
> enough.
>
> -----Original Message-----
> From: J.E.C.Brisland [mailto:j.e.c.brisland@open.ac.uk]
> Sent: 05 November 2010 11:57
> To: 'JMeter Users List'
> Subject: Best way to build a test...
>
> Hi All,
>
> Basically what I want to do is reproduce the average load our server has on
> a daily basis onto our test server.
>
> We have a number of different page types on our server, and each of these
> types gets hit a varying amount.
>
> Take for example blog views, posts and comments.
>
> Lets say these are our numbers
>
> Per second
> 0.5 users log in
> 3.4 blog views
> 0.2 blog posts
> 0.002 blog comments
>
> What would be the best way to reproduce this load using jMeter?
>
> I have tests running, but I'm struggling to reproduce the loads I want.
>
> --
> The Open University is incorporated by Royal Charter (RC 000391), an exempt
> charity in England & Wales and a charity registered in Scotland (SC 038302).
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
>
>
--
Atenciosamente
Regards
Evandro Grezeli de B. Neves
RE: Best way to build a test...
Posted by "J.E.C.Brisland" <j....@open.ac.uk>.
Note: I also have a whole bunch of other things I want to add, for example
Per second
4.8 forum views
0.2 forum posts
0.9 general page views
Etc..
The concurrency of these requests is not known, so I just want to randomise it a little, as long as I can get a general load per second it doesn't really matter (to me anyways) about overlap. If it's random it will be good enough.
-----Original Message-----
From: J.E.C.Brisland [mailto:j.e.c.brisland@open.ac.uk]
Sent: 05 November 2010 11:57
To: 'JMeter Users List'
Subject: Best way to build a test...
Hi All,
Basically what I want to do is reproduce the average load our server has on a daily basis onto our test server.
We have a number of different page types on our server, and each of these types gets hit a varying amount.
Take for example blog views, posts and comments.
Lets say these are our numbers
Per second
0.5 users log in
3.4 blog views
0.2 blog posts
0.002 blog comments
What would be the best way to reproduce this load using jMeter?
I have tests running, but I'm struggling to reproduce the loads I want.
--
The Open University is incorporated by Royal Charter (RC 000391), an exempt charity in England & Wales and a charity registered in Scotland (SC 038302).
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org
Re: Best way to build a test...
Posted by Carl Shaulis <cs...@homeaway.com>.
My recommendation would be to convert your production access log file into a
flat file of http requests, then use JMeter to replay the flat file (.csv).
You will need to execute a ramp to determine when the transactions per
second mimic the load of production.
The assumption unfortunately is that you have consistent data between the
two environments.
Alternatively you can create the 4 scripts described below and execute them
in proportionate number of active of threads.
Good luck!
Carl
On 11/5/10 6:57 AM, "J.E.C.Brisland" <j....@open.ac.uk> wrote:
> Hi All,
>
> Basically what I want to do is reproduce the average load our server has on a
> daily basis onto our test server.
>
> We have a number of different page types on our server, and each of these
> types gets hit a varying amount.
>
> Take for example blog views, posts and comments.
>
> Lets say these are our numbers
>
> Per second
> 0.5 users log in
> 3.4 blog views
> 0.2 blog posts
> 0.002 blog comments
>
> What would be the best way to reproduce this load using jMeter?
>
> I have tests running, but I'm struggling to reproduce the loads I want.
---------------------------------------------------------------------
To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jmeter-user-help@jakarta.apache.org