well since this question was re-re-asked (again :) ), here goes nothing:
how practical with you consider using the __javaScript() function in
this type of situations? (or any other way to process client side
actions when they are strictly needed)
( I am thinking of situations were the
an applet can do(in terms of client-server
communication)
,
you can do in Java/BSH.
regards
deepak
On Thu, Jul 16, 2009 at 7:29 AM, Adrian Speteanu asp.ad...@gmail.com
wrote:
well since this question was re-re-asked (again :) ), here goes
nothing:
how practical with you
There is a workaround to this:
- depending on the way you organize your test plan tree, your
listeners record only samples from there level down (inclusive) - I
know I read this somewhere, but could not find a reference link for
you.
example:
1. login
|-[ simple controller ]-
|- 2. user
It doesn't seem related to the server or the application tested. In
the logs of JMeter nothing is written during the test (i found only
start and stop threads).
Any ideas on why it happens / how to avoid it?
[ the error message ]
java.net.BindException: Address already in use: connect
at
this point of
view (I assume).
On Thu, Jul 23, 2009 at 5:29 PM, sebbseb...@gmail.com wrote:
On 23/07/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
It doesn't seem related to the server or the application tested. In
the logs of JMeter nothing is written during the test (i found only
start
, Adrian Speteanu asp.ad...@gmail.com wrote:
Thanks for the idea.
I'll try the HttpClient sampler instead.
Perhaps add Connection close to some of the samplers.
Is it possible to do this with the sampler? How?
Sorry, that was not clear - I meant clear the Keep-alive check-box.
About the RAM
Hi Toni,
Take a look at this as well:
http://wiki.apache.org/jakarta-jmeter/LogAnalysis if you haven't
already.
Basically if you want to write xml, to save more data , you can find
there a method to use stylesheets and convert the data into a table.
On Fri, Jul 24, 2009 at 3:16 PM,
I am the one that have asked this question before.
I haven't got the time to retest with the recommendations - to see if
they solve anything.
I am also unable to reproduce the bindException every time I try,
however it took me only 5 test threads to encounter this issue. So I
was also think that
Guys,
As I mentioned before, I've reproduced this problem with only 10
threads, testing a server that wasn't on the local machine and was not
being used for anything else than the test.
Therefore, we can eliminate the fact that it takes only high loads to
reproduce the issue.
I test with jmeter
On Thu, Jul 30, 2009 at 7:26 PM, Deepak Shettyshet...@gmail.com wrote:
hi
ok, I assume you've tried out the other things recommended by sebb (and i
think emailed by adrian).
Do you get any errors on tomcat when you start getting this error on the
JMeter client
Can you monitor the sockets on
Hi guys,
I am trying all sorts of methods to generate better logs and better
ways to interpret the data obtained.
In jmeter.properties, there is this very useful piece of information:
# If the filename contains paired single-quotes, then the name is processed
# as a SimpleDateFormat format
apache tomcat can use some logkit to rotate it's log files.
is it not possible to add such existing utilities to JMeter?
On Tue, Aug 4, 2009 at 3:45 PM, sebbseb...@gmail.com wrote:
On 04/08/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
Hi guys,
I am trying all sorts of methods
a-haaa. too bad, that is out of the question.
it would have been nice ( IF it would be more simple than what you say...)
On Wed, Aug 5, 2009 at 3:15 PM, sebbseb...@gmail.com wrote:
On 05/08/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
apache tomcat can use some logkit to rotate it's log
heap memory refers to ram memory.
the fact that this happens for 1000 users and not for 200 users means
that your machine can cope with only 200 users for this test plan,
thus you need to distribute the load on 5 machines.
check this out:
sorry about that, I wasn't clear. you need to establish how many users
are enough on your machine. and to reach 2500 users as much as you
need, you have to distribute the test-load.
On Fri, Aug 7, 2009 at 12:44 PM, Adrian Speteanuasp.ad...@gmail.com wrote:
heap memory refers to ram memory.
On Fri, Aug 7, 2009 at 12:50 AM, wsmith84tobymartind...@gmail.com wrote:
I have a CSV file I'm reading in delay times from for start times for a soap
message sampler
basically I'm trying to first recreate actual usage of a web service by
setting
up a csv file with a variable called
Hi,
For this request:
loginRequest?username=xuserpass=yticket=z_event=submit
In the result tree I get:
-home page
|- loginRequest? all parameters as above
- set-cookies headers
|- redirect1
- the cookies above were intrepreted and a new unique value is
generated to validate the
yes, but it should work for him to just use the cookie manager. that
is the best solution to deal with jsessionid.
or he can just make the request as deepak suggested:
.../something.jsp;${extracted_jsessionid}?variable=xand_so_on_if_required
(directly in the http sampler)
also, he said he found
select. There's no way to select a
particular child sample.
regars
deepak
On Thu, Aug 13, 2009 at 1:02 AM, Adrian Speteanu asp.ad...@gmail.comwrote:
Hi,
For this request:
loginRequest?username=xuserpass=yticket=z_event=submit
In the result tree I get:
-home page
Hi Andrey,
I've encountered you're situation as well in different circumstances:
I have to test something like
http://localServer1:port1/folder1/various actions i want to test
However, the programmers change localServer1 and folder1 ALL THE
TIME due to the various branches of the
, Andrey Simonovau...@bk.ru wrote:
Adrian,
Thank you for you answer.
I'm looking for a way not to specify the default folder at all, even without
variable.
Adrian Speteanu wrote:
Hi Andrey,
I've encountered you're situation as well in different circumstances:
I have to test something
On Wed, Aug 19, 2009 at 2:05 PM, sebbseb...@gmail.com wrote:
On 19/08/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
Well, in that case, the only solution I can think of is to modify the
code and customize the behaviour of the http defaults.
That's the best long-term solution.
I thought
The solution with multiple test plans is definitely the simplest.
You can run it also from GUI mode, but you need to machines (you can
use the schedule option from the thread gui element to start testing
simultaneously on both machines).
Another question about parallel execution:
How to make the
Hi Andrey,
I know this might be confusing when first starting to use jmeter, so I
think a beginners' advice might be of more use:
From your question I assume you want to check if an action doesn't
receive a slow answer and fail it if so. Use this:
- sampler
|- duration assertion
Assertions
Hi guys, I was experimenting with an idea that doesn't require coding
and found a little bug:
Make a test plan like this
thread group (threads = 1, ramp = 1, loops = 1)
+loop controller (5 times)
\- module controller that refferences to the random controller bellow
+recording controller
\-
\- random controller (under which samples are recorded with the proxy
server)
\- http sampler 1
\- http sampler 2
On Thu, Aug 27, 2009 at 9:26 AM, Adrian Speteanu asp.ad...@gmail.comwrote:
Hi guys, I was experimenting with an idea that doesn't require coding
and found a little bug
)
Module Controller --- Thread Group 2 --Recording -- Simple
Thread Group 2 (Disabled)
Recording Controller
Simple Controller
Req 1
Req 2
Result = 5 requests (Req1,Req2,Req2,Req2,Req1)
regards
deepak
On Thu, Aug 27, 2009 at 10:18 AM, Adrian Speteanu
Hi Deepak,
What I meant is that in your test plan example there wasn't a Random
Controller anymore.
I've tested like this and it works for me too:
Thread Group 1 (1 thread 1 Loop)
Loop Controller ( 5)
Module Controller --- Thread Group 2 --Recording -- Simple
Thread Group 2 (Disabled)
it is also good to get the request you make in jmeter (copy it from
the Request tab in the Results tree) and use it directly in the
browser, see if it works there.
check if the login is successful in the browser like you do it in
jmeter first. if it works in the browser and not in jmeter, then
true, you can use either method for what you said you need, but in
this case, saving the file on the test machine will significantly
increase the stress on the test environment (quality image files mean
lots of space and that means disk usage).
if you run the test with fewer requests and see that
Not mentioned: use the workbench to migrate data from elements saved
to the working test plan.
Example: you save a thread group or a controller that performs a
test-case-a1. Merge them in the workbench then copy paste what you
need to the new test plan.
In order to be able to migrate settings
if you run the jmeterw.bat you also won't be sending the extra
parameters to JMeter. the tweaks in jmeter.bat are useful and don't
know any other way to get around this.
is there any difference between running java.exe and javaw.exe? (i
always used javaw)
On Fri, Sep 4, 2009 at 3:03 PM,
On Fri, Sep 4, 2009 at 4:28 PM, sebbseb...@gmail.com wrote:
On 04/09/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
if you run the jmeterw.bat you also won't be sending the extra
parameters to JMeter. the tweaks in jmeter.bat are useful and don't
know any other way to get around
well, it shouldn't stay exactly the same, but it could be
mathematically possible to get a similar throughtput, though very
unlikely. so it matters also what timers you use, how fast the server
responds to these requests etc etc.
depending on your test plan and the timers you set, it is possible
there are other libs that are difficult to find, for example the links
from here :
http://jakarta.apache.org/jmeter/usermanual/get-started.html#get_started
are broken (all but beanshell.org - but that jar comes with the
default instalation, i think)
On Wed, Sep 9, 2009 at 11:41 AM,
/products/jms/docs.html
- this might be useful for some.
Also, after the web service specific test plan link
(http://jakarta.apache.org/jmeter/usermanual/build-ws-test-plan.html
there is a bullet, but no text there.
On Wed, Sep 9, 2009 at 12:45 PM, sebbseb...@gmail.com wrote:
On 09/09/2009, Adrian
...@gmail.com wrote:
On 09/09/2009, Adrian Speteanu asp.ad...@gmail.com wrote:
there are other libs that are difficult to find, for example the links
from here :
http://jakarta.apache.org/jmeter/usermanual/get-started.html#get_started
are broken (all but beanshell.org - but that jar comes
3 x 90 = 270
you set the timer to 280 threads! normally, it shouldn't execute
anything because there will never be 280 threads ready to execute the
next step.
in fact, since there are three different threads, shouldn't the timer
be set to 90? I actually don't know this, I never thought of using
Trial and error. Try different combination of parameters and see for
your test plan what happens, that is how you find out this.
On Thu, Sep 10, 2009 at 11:55 AM, Hembha, Ambanna
ambanna.hem...@logica.com wrote:
Hi,
To achieve the 100 concurrent users login+search test as advised below, how
I think that it might be something about connections not getting
closed properly (this has been discused). I got to the point that i
ran out of sockets because of this, but in your case, since tomcat
only accepts a little number of connections, it is easy to run out of
open connections and you get
http://jakarta.apache.org/jmeter/usermanual/listeners.html#sample_configuration
in the gui element, you set the file you want the information to be written to.
you make the configurations you see in the link above (uncheck Save
as XML so you save as CSV). also, check save column names (CSV) -
you
you should take a look at md5 assertions (might be what you are looking for).
On Mon, Sep 21, 2009 at 9:21 PM, Deepak Shetty shet...@gmail.com wrote:
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Save_Responses_to_a_file
regards
deepak
On Mon, Sep 21, 2009 at 11:19
you have controllers like the once only controller and the throughput
controller (the latter allows to define percentages or a number of
executions), assuming this is what you were asking.
On Sat, Oct 10, 2009 at 4:58 AM, sebb seb...@gmail.com wrote:
On 09/10/2009, b.ram...@eventim.de
there should be tools for the operating system that server uses and the
net-admin should / could help you with what you need.
On Wed, Oct 14, 2009 at 1:17 PM, sebb seb...@gmail.com wrote:
n 14/10/2009, Jatin Davey jasho...@cisco.com wrote:
Hi All
Is it possible to do the following with
Hi,
When you load your test plan again merge the proxy you have save so you can
keep it with the loaded test plan. It's a step more than what you expected,
but it makes it possible.
From what I know,, the workbench is mostly for copy-paste actions and for
the non-test elements. I find it
- /homepage, -2- /otherpage, etc and they get hard to
find once you have more of them. I have to import the log files into xls and
sort them while the test is still in progress.
have an easy day.
On Wed, Nov 4, 2009 at 4:10 PM, Nikolay_Miroshnichenko nikb...@gmail.comwrote:
Adrian Speteanu wrote
On Tue, Nov 17, 2009 at 3:29 PM, cmrz c...@yahoo.com wrote:
Hi,
I have a couple of questions . . .
1) I am running Jmeter via command prompt. I am executing the login
script.
Test pulls up the login page, enters user name and password, clicks the
login button and logs the user in.
hi,
shouldn't the checkbox Retrieve all embeded resources help download all
required resources? (including imagines, .js files, xml files, etc)
On Wed, Nov 11, 2009 at 11:49 PM, Deepak Shetty shet...@gmail.com wrote:
hi
I dont think you can do this directly/ dynamically. Statically if you
if the id is made only of digits, I usually use (\d+) to extract all of
them; example:
\item_id\: \(\d+)\
the (\d+) example should work especially when the digits are surrounded by
other pieces of text that you don't want to extract (if the expression is
too general it will extract a lot of
i'm a bit rookie with the regex, but this one seems pretty easy, use this
expression:
ref_name: extractedSessionId
regex:=(.+)
template: $1$
match: 1
default: whatever
if you call ${extractedSessionId} you will receive your sessionid including
the characters from your
On 2/12/10, Carl Shaulis cshau...@homeaway.com wrote:
I would suggest record each path as a thread group, then combine all thread
groups into a single script. Set thread group 1 to run 50% of the total
threads, thread group 2 25% etc...
for more complex test configurations, it can also be
hi guys,
i had the same issue about an year ago. i had to simulate with one
thread the user's actions and with another some requests made by the
application regularly, in the background (recording the requests for
certain use-cases was just not enough).
so, i have created two thread groups and
manager
(using beanshell) or append URL.
Never tried it dont know what issues you might run into.
regards
deepak
On Thu, Mar 18, 2010 at 6:31 AM, Adrian Speteanu asp.ad...@gmail.comwrote:
hi guys,
i had the same issue about an year ago. i had to simulate with one
thread the user's actions
hi everybody,
sorry to reopen such an old issue, but i have remembered some
discussion threads about closed sockets and other socket issues
(mostly when testing from a Windows station) and thought to bring
something up: http://half-open.com/
on some versions of Windows, the tcpip.sys is not
also a good question is: does the number of threads remain 0? (if this is
the case, it is usually a test plan configuration issue - for example you
use a schedule that is in the past, or you used variables for thread
configuration and one of them is not set, or 0...)
On Fri, May 21, 2010 at 2:27
hi,
@deepak: the [^] regex doesn't seem to work for me as you have mentioned it
(i'm on a windows station).
[ details ] i have this in the text response:
name=lt value=
_c12F59EBD-2BE6-A12A-B0B8-BCA9A6F71683_k1D7CB867-93BB-4F79-F306-4A61835AB21C
(which is the ticket required by CAS for login -
You could also read Emily H. Halili - Apache JMeter, but you should know
that it has only a little more form the usermanual (
http://jakarta.apache.org/jmeter/usermanual/index.html), but it is more
user friendly, explains things in the logical order.
I've read it after going through various
I prefer alphabetical ordering myself, makes sense.
But it's not that a big of an issue. You don't use JMeter for general
purposes, but for specific tasks. So you can always hide the samplers you
don't really need / use (from jmeter.properties). My sampler menu has only
6-7 entries (and probably
you have to increase the virtual memory size, if on a windows station and a
defragmenting of the OS disk-partition might be useful for the space to
provide room for the JVM requirements. I've encountered a similar JVM crash
during load tests and after some research found some opinions / comments
hi,
configured the jmeter-maven plugin. jmeter tests are executed and produce
the output log configured in the test plan.
however, in case errors are reported in the log, maven still reports a
successful build. is there something that can be configured in the test plan
so maven sees an error was
Hi,
Depends what you understand by fine :). It would be easier with an example
of test-plan and what do you expect to happen.
- sample1
- sample2
- sample3
- Timer
- results tree
If timer is constant timer, than they will be executed in the same order.
If timer is gaussian random timer, than it
Just a suggestion, it still won't be easy:
- test machines with synchronized clocks (look at about time in case of
winodws, similar tools in case of other OS);
- synchronizing of the threads occurs at certain intervals;
- use beanshell to control this...
itl won't be pretty, though. Maybe
Or you can take a look at jMeter's Counter (Config Element).
It usually does the job.
On Mon, Oct 4, 2010 at 8:11 AM, kiransiri
sirigiri_kiranku...@yahoo.co.inwrote:
hi Din,
use the function ${__counter(1,)} in the value where you want to increase
the number
Regards,
Sree Kiran
--
Hi guys,
These threads are juicy... I would like to add 2 things:
- first, if you test embeded resources and you get bad results - what then?
There isn't much that can be improved because it's handled by the container
which you take it as it is (not more then what tools like YSlow already tell
Hi,
Also try checking Include group name in label? checkbox (down in the
Aggregate Report GUI) and run the test again. See if this is what you need.
Regards,
Adrian S
On Mon, Nov 1, 2010 at 10:22 AM, SanderW
san...@performancearchitecten.nlwrote:
Hi Rashmi,
I don't know if I understand
Hello,
Had some problems with setting jmeter.save.saveservice.default_delimiter to
\t (just uncommented the line in jmeter.properties).
jre = 1.6.0_20
jmeter.properties
# For use with Comma-separated value (CSV) files or other formats
# where the fields' values are separated by specified
done
On Wed, Nov 3, 2010 at 5:07 PM, sebb seb...@gmail.com wrote:
On 3 November 2010 14:57, Adrian Speteanu asp.ad...@gmail.com wrote:
Hello,
Had some problems with setting jmeter.save.saveservice.default_delimiter
to
\t (just uncommented the line in jmeter.properties).
jre
Hello,
Can you think of a way to assert that something is matched in document on
the exact x match?
I have:
1. variableToMatch - its numeric
2. samplerToAssert - response matches ${variableToMatch} pattern several
times, an assertion without regex will also pass easily, but it is more
important
hello,
is jmeter POST request being encoded (in request tab of view results tree)?
it's not very clear from your mail, my understanding is that the later
response returns encoded.
if jmeter sends the request as you have pasted it bellow, when you POST the
data initially, then it means that the
to
context, so some incorrect substitutions may be made.
rgds,
Frank.
- Original Message -
From: Adrian Speteanu asp.ad...@gmail.com
To: JMeter Users List jmeter-user@jakarta.apache.org
Sent: Monday, November 8, 2010 2:02:19 PM
Subject: Re: prevent POST encode
Hello,
Hope these help (personally I read the last link first, some time ago, just
to have some clue on the parameters used in jmeter script, as they come out
of the box: .bat for windows or jmeter on linux). However they still should
be useful.
Hello,
Need some advice on configuring jmeter shell script to run better on linux
(debian/ubuntu in my case).
Uncommenting these parameters in jmeter script has adverse affects:
# SURVIVOR=-XX:SurvivorRatio=8 -XX:TargetSurvivorRatio=20%
# EVACUATION=-XX:MaxLiveObjectEvacuationRatio=20%
The JVM
In version 2.4 you have one very useful difference in HTTP Request HTTP
Client : IP spoofing.
On Mon, Nov 29, 2010 at 11:52 PM, Rob Schroder r...@manilla.com wrote:
I have not. And, apologies for not doing so before posting the question!
Thanks for the pointer.
On 11/29/10 12:36 PM, sebb
Hi guys,
In addition:
a. If you suspect that the server behaves differently when receiving
requests from the same IP, you could double check this using the new feature
in version 2.4, ip spoofing (available only for the Http Request Http
Client sampler). Note that some network admins might not
that being said, you can explain throughput as the speed at which the
server responds to all requests taken into account.
On Wed, Jan 5, 2011 at 9:57 PM, Deepak Shetty shet...@gmail.com wrote:
You arent factoring in number of threads
Lets say you have 1 thread makes 10 requests each takes 3
Hello,
I cannot open from GUI a .jtl file saved in CSV format (not even with the
same listener used to save it to disk).
jmeter 2.4
This is in jmeter.log:
2011/01/10 15:20:20 WARN - jmeter.save.CSVSaveService: Error parsing field
'timeStamp' at line 1 java.text.ParseException: Unparseable
Hi,
Problem solved.
On Mon, Jan 10, 2011 at 7:58 PM, sebb seb...@gmail.com wrote:
On 10 January 2011 13:41, Adrian Speteanu asp.ad...@gmail.com wrote:
Hello,
I cannot open from GUI a .jtl file saved in CSV format (not even with the
same listener used to save it to disk).
jmeter 2.4
The jdk comes bundled with the JConsole wich is meant for java profiling
(you can monitor java processes real-time). There are a lot of similar tools
out there (for free or payed), you can google it around, see which one has
what you need.
On Thu, Jan 13, 2011 at 8:41 AM, Abhishek Kona
Plugin).
-Abhishek Kona
On 13/01/11 3:04 PM, Adrian Speteanu wrote:
The jdk comes bundled with the JConsole wich is meant for java profiling
(you can monitor java processes real-time). There are a lot of similar
tools
out there (for free or payed), you can google it around, see which one has
On 13/01/11 8:53 PM, Adrian Speteanu wrote:
Hi,
JMeter is for benchmarking other applications, so you have the listeners
that give you the statistics you need to do that. So, not sure how / what
you want to do, but have a look at this:
http://jakarta.apache.org/jmeter/usermanual/build
Hello everybody,
Need help / suggestions about test idea. Bellow you have the test plan ( the
while controller is setup just like in the CSV example that comes with
jmeter):
thread group [ loop = forever ]
+ While controller
[ While condition : ${__jexl(${somethingFromCSV} != EOF)} ]
On Tue, Jan 18, 2011 at 7:03 PM, sebb seb...@gmail.com wrote:
On 18 January 2011 16:18, Adrian Speteanu asp.ad...@gmail.com wrote:
Hello everybody,
Need help / suggestions about test idea. Bellow you have the test plan (
the
while controller is setup just like in the CSV example
@10 threads, I don't think it makes a difference over time if you are using
GUI mode, even for long times.
1. Have you tried with -Xmx1024? too see if it works at that level without
memory exceptions or you raised the size because of memory problems in the
first place?
2. Edit the jmeter script
Hey,
That sounds complicated. This is how I did it:
Problem
- application requires lots of .xml documents for load test (more than I
can generate prior to test, or even store for that matter)
- I wanted to have access and modify dynamically the content sent in those
xml
Solution
1. use Soap
Hello,
I started learning this tool: http://www.gnuplot.info/
because I need to run tests on remote linux machines anyway and because it
has its own scripting support that can be executed at the end of test
automatically and in the morning you just get the test results and some .png
files with
| www.mphasis.com
- Print this mail, only if you need to...
-Original Message-
From: Adrian Speteanu [mailto:asp.ad...@gmail.com]
Sent: Wednesday, February 16, 2011 9:32 AM
To: JMeter Users List
Subject: Re: Jmeter 64 bit Out of memory exception
@10 threads, I don't think it makes
This might cause a problem:
Don't use functional mode - if this is checked in the test plan GUI
element than it might cause a problem with using lots of memory.
The exception you have mentioned is not related to something of jmeter's I
think (though not completely sure), but rather related to
Cannot creative native thread is also related to system software
limitations on desktop distributions (that can usually be tweaked). On
windows 7 however, java runs nicely... I also got some pretty stable results
at 4500 threads with some basic tests just by running jmeter without any JVM
I think there would be another convienient approach, see bellow:
On Thu, Mar 31, 2011 at 7:36 AM, Sonam Chauhan son...@gmail.com wrote:
I use a tool that runs JMeter instances in parallel on one machine. When I
upgraded from JMeter 2.3 to JMeter 2.4, I hit this error running 2 JMeter
Note that ulimit -n gives the limit per user, you should also check the
limitation per system as well.
Any socket exception are obviously related to the OS and / or network.
On linux, this problem in particular is related to the fact that linux
treats open sockets as open file descriptors. And
ulimit -n is usually at 1024 on latest debian / ubuntu distributions. that
is a little too small for web-applications performance tests. it can be
increased to 10-30,000, though I agree a check on the test should be done.
On Thu, Apr 7, 2011 at 9:57 PM, Deepak Shetty shet...@gmail.com wrote:
Hello sebb,
Will adjusting the headers work for this request by using the Header
Manager? Last time I tried, the headers weren't sent (I assumed the header
manager only works for Http Samplers).
Regards,
Adrian
On Wed, Apr 13, 2011 at 2:41 PM, sebb seb...@gmail.com wrote:
On 13 April 2011
My bad, it does work.
On Thu, Apr 14, 2011 at 6:35 PM, sebb seb...@gmail.com wrote:
On 14 April 2011 14:59, Adrian Speteanu asp.ad...@gmail.com wrote:
Hello sebb,
Will adjusting the headers work for this request by using the Header
Manager? Last time I tried, the headers weren't sent (I
Download java JRE - all files keep directory structure, place it in the
location of your choosing and then edit the jmeter file located in
jmeter's bin directory. At the end of the file you will have the command
line used to run jmeter something like [java argument -jar
AparcheJmeter.jar
On Mon, May 9, 2011 at 1:07 PM, apc a...@apc.kg wrote:
Maybe it will be more suitable to force all agents to work for the same
time
period and don't finish until then?
I concur - this can be done by using the scheduler in the thread group GUI
config.
--
View this message in context:
rodolfo.k...@intel.comwrote:
Then I would configure threads to never end and add a scheduler?
Thanks.
Rodolfo
-Original Message-
From: Adrian Speteanu [mailto:asp.ad...@gmail.com]
Sent: Monday, May 09, 2011 2:30 PM
To: JMeter Users List
Subject: Re: Throughput with many jmeter
Hi Bruce,
I've given a quick look on your extensions and* JmeterThreadGlobal seems
interesting enough.* The idea behind it was discussed one or two years ago,
people always asked about running threads inside other threads or sharing
between threads. Myself inluded: I needed at some point to use
Hey,
Use the gaussian random timer, its the best.
It has the advantage of giving random timeouts with a higher probability for
a value in the middle of the range (at the end if you plot all numbers
returned on the rate of occurrence by it you'll have a gaussian
distribution). so if you know that
Hi,
This doesn't sound right, are you sure? Response time is calculated from the
moment the request was sent to the moment when the last bit of the response
was received. So it can't include wait times. If your sampler would do some
processing (for example the beanshel sampler), the response time
Hello all,
John, Sorry to bargin on the discussion mid-way, from your first description
of the problem, I wonder if you are running JMeter by using the jmeter
script (for linux) or jmeter.bat (for windows).
I have a graph of a benchmark that emphasizes exactly what you have
described in your
1 - 100 of 136 matches
Mail list logo
