rk access. In my case I run
Firefox off a UNIX server to record my tests.
--
Bruce Ide
flyingrhenqu...@gmail.com
app. Turns out they only validate
it in Javascript. I can happily drop a 2 gigabyte file into the database by
calling the server API with a gigantic file. You could probably find a ton
of those types of bugs in your typical AJAX application.
--
Bruce Ide
flyingrhenqu...@gmail.com
You probably need to configure your browser to accept the Jmeter SSL
certificate.
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#HTTP_Proxy_Server
Read the bit on recording HTTPS.
--
Bruce Ide
flyingrhenqu...@gmail.com
kly run out of memory.
My one test that does something like that uses two listeners that write
successful and failed requests to disk. I can just fire it up and if I see
any failed requests then I know my test failed.
--
Bruce Ide
flyingrhenqu...@gmail.com
" to jmeter when you invoke
it, and then set the start and end times in the scheduler to
${__P(START_TIME,)} and ${__P(END_TIME,)}, but I'm not positive if that
would work or not.
--
Bruce Ide
flyingrhenqu...@gmail.com
ding controller
to recognize messages in that protocol and use the appropriate sampler for
them.
--
Bruce Ide
flyingrhenqu...@gmail.com
You do this in a User Defined Variables Configuration Element, or in the
test plan element configuration page. Either one will work.
--
Bruce Ide
flyingrhenqu...@gmail.com
rk to the
local exceptions and if the application you're attempting to test is on the
internal network with that exception set up, jmeter won't record any of
those pages.
--
Bruce Ide
flyingrhenqu...@gmail.com
il. I don't know if it does much other than that. One of my co-workers is
fond of using them but I haven't used them all that much in my test. Do any
further statements in the controller execute once one fails? I could see it
being useful if it prevented any from executing after that point.
just replace
your static variable ID with the jmeter variable you're using in your
ForEach loop.
--
Bruce Ide
flyingrhenqu...@gmail.com
Couldn't you just use a JDBC config element and sampler?
--
Bruce Ide
flyingrhenqu...@gmail.com
Thanks for the info
On Tue, Sep 27, 2011 at 11:51 AM, sebb wrote:
> On 27 September 2011 19:04, Bruce Wobbe wrote:
> > Hi All,
> > is there a way to get the sample number produced in the jtl? I've tried
> >
> That's the number of samples in
ons I use in the webapp.
Once you get your login test recoded, play it back and make sure it works.
Then you can tinker with it.
--
Bruce Ide
flyingrhenqu...@gmail.com
e for example sample 77 is slow
I want to be able to go into the jtl and quickly find sample 77
Bruce
parameter with, in your string.
I apologize if I didn't explain that very well, I haven't had my coffee yet.
If it didn't make sense to you, I'll try to dig up some additional
information about it from my research on the subject for you.
--
Bruce Ide
flyingrhenqu...@gmail.com
dons to see if
they work for you. They're licensed under the same Apache license Jmeter is,
and the source code and Jmeter 2.4 jars are available on github. Keep in
mind that when you do something like this it makes for an extremely
complicated test that no one else will understand. My test worked great, but
no one else understands it.
--
Bruce Ide
flyingrhenqu...@gmail.com
ystem I used to work on
would store 30+ megabyte Strings on a regular basis. You definitely don't
want to create any more copies of those than you have to!
--
Bruce Ide
flyingrhenqu...@gmail.com
You don't need to use ""+ to convert bshFoo into a String! It's already a
String! And using ""+ to convert something to a String is the wrong way to
do that, anyway!
God!
;-)
--
Bruce Ide
flyingrhenqu...@gmail.com
Yep, thanks I figured this out right after I sent the email
Bruce
On Thu, Sep 22, 2011 at 2:16 PM, Bruce Ide wrote:
> To pull java properties into jmeter variables you have to use the __P
> function. Try making a user defined variable config element with TestServer
> defined, and make
property isn't defined, you can put it in there.
--
Bruce Ide
flyingrhenqu...@gmail.com
class
org.apache.jmeter.testelement.property.StringProperty: ${TestServer}
2011/09/22 13:50:08 DEBUG - jmeter.engine.util.ValueReplacer: Replacement
result: ${TestServer}
Am I doing this wrong? What is the issue here?
Thanks
Bruce
good idea
On Tue, Sep 20, 2011 at 1:51 PM, Deepak Shetty wrote:
> hi
> sure - but its talso the most efficient way for Jmeter. we usually have an
> additional step in ANT to do all these tasks - before kicking off jmeter
>
> regards
> deepak
>
> On Tue, Sep 20, 2011
Thanks Deepak,
I went with the aggregator approach of step 3 ,it adds some overhead to my
tests but works pretty simly
Bruce
On Mon, Sep 19, 2011 at 9:22 PM, Deepak Shetty wrote:
> The problem is if you have Recycle as true , you never get EOF, if you have
> recycle as false then the inn
7 calls to HTPRequest, I get
3
Are nested while loops not supported in this fashion?
Thanks
Bruce
Could you use a synchronizing timer to insure your threads run more
consistently? I find them to be underrated, and they're worth taking a look
at if you haven't yet.
--
Bruce Ide
flyingrhenqu...@gmail.com
I would swear that I've used the SOAP request for that in the past. My
service wanted a post in XML format, but it doesn't seem to provide a WSDL
or anything I'd normally associate with SOAP. If your service wants some
XML, I'd suggest giving that a try.
--
Bruce Ide
flyingrhenqu...@gmail.com
to be a
bit... unorthodox... though.
--
Bruce Ide
flyingrhenqu...@gmail.com
There's an open source NTLM proxy, cntlm, which might let you work around
it. It wasn't terribly difficult to configure when I looked at it. Works on
UNIX too. If you're really hard up...
--
Bruce Ide
flyingrhenqu...@gmail.com
er iterator elements too, but consult the user's manual
if you haven't used them before, as their syntax may not be what you expect.
--
Bruce Ide
flyingrhenqu...@gmail.com
ounter));
You could probably do this in an __eval or __javascript or something, but I
don't know if it really buys you any more or less trying to do it that way.
I tend to prefer BSF samplers (With Groovy) for being more readiable. To a
java programmer anyway...
--
Bruce Ide
flyingrhenqu...@gmail.com
t's
due to Javascript you'll just have to compute a constant execution time
that's added by the Javascript. If this is the case you'll probably want to
compute different ones for each browser, too, since Chrome handles
Javascript a lot faster than Firefox does.
--
Bruce Ide
flyingrhenqu...@gmail.com
It's fine, just ignore it.
--
Bruce Ide
flyingrhenqu...@gmail.com
t the
systemwide environment settings.
--
Bruce Ide
flyingrhenqu...@gmail.com
iew results tree listener (That you can disable when not
debugging the test) and one or more debug samplers at various points in the
test. That makes it very easy to see what the test is doing, and what
variables are being created at various points in the test.
--
Bruce Ide
flyingrhenqu...@gmail.com
memorized, but
you should be able to turn up some good examples if you google around on
"Jmeter BSF Preprocessor".
--
Bruce Ide
flyingrhenqu...@gmail.com
change
something in samplevars.jmx, you need to save BOTH samplevar.jmx and
testSampleVars.jmx, so that the include controller refreshes its tree
correctly.
--
Bruce Ide
flyingrhenqu...@gmail.com
-
To unsubscribe, e-mail: jmet
I've always wondered why "use tidy" isn't on by default. My Xpath usually
looks good to me, but it doesn't ever actually work unless I actually select
that option. It's like having an option, "Make this component actually work"
that is disabled by default.
--
Bruce Ide
flyingrhenqu...@gmail.com
ways prefer the xpath extractor over the regexp extractor when I can use
it, as it takes much less fiddling to make it work. Usually with the xpath
extractor I just need to remember to click "use tidy" and I'm good.
--
Bruce Ide
flyingrhenqu...@gmail.com
eate an array if you specify a default
value like that and it assigns the default. It'll put "5000" in keyval, not
keyval_1.
--
Bruce Ide
flyingrhenqu...@gmail.com
r
specific environment. You could probably set it up to pull its data from
java properties files too.
--
Bruce Ide
flyingrhenqu...@gmail.com
D around. Doing things in separate
sessions is really the only reason I'd run multiple threads, and I wrote the
aforementioned plugin to support something like that.
--
Bruce Ide
flyingrhenqu...@gmail.com
If you click on your "Test Plan" node, there's an option to "Run Thread
Groups Consecutively" which is probably what you want. If you need more
intra-thread control, you might need to look into some of the timers. Give
that a try first though.
--
Bruce Ide
flyingrhenqu...@gmail.com
ou're looking for to a variable with an
xpath or regex post-processor.
You can use a cookie manager for session management.
--
Bruce Ide
flyingrhenqu...@gmail.com
e's not one in there. You're probably sending the old session ID
somehow, and it's just a matter of finding the old one and removing it so
that the cookie manager can do its job.
--
Bruce Ide
flyingrhenqu...@gmail.com
Did you record the test previously using the http proxy? I bet your test is
trying to use the old session ID that probably timed out a couple of days
ago.
I think you can just slap a HTTP cookie manager config element at the start
of the test and that should fix it.
--
Bruce Ide
flyingrhenqu
Yikes, every one? I wonder if it would change if you put that code in an
http header manager config element up at the top of the test? That would be
outlandishly cool if that would actually work...
--
Bruce Ide
flyingrhenqu...@gmail.com
y work pretty well! It's one of those things where it seems like
you're going to make mistakes and just have to evolve good practices to deal
with the issues based on that. If you read this entire mail and decide to
use the include controller, I probably just saved you a week or two of that
sort of pain and suffering.
--
Bruce Ide
flyingrhenqu...@gmail.com
be pretty
spiffy. Of course, I probably shouldn't be doing things like that with
jmeter anyway.
--
Bruce Ide
flyingrhenqu...@gmail.com
This may or may not be what you want.
The second one gives you the class name of the object. That's pretty cool.
I'm not sure if you can get at the sampler's generic name though. If you
just want "Java Request" no matter what the Name field is, I don't know how
y
Is your test able to shut down in non-GUI mode if it just exits normally?
There have been problems in the past with third-party addons preventing
jmeter from exiting while in non-GUI mode.
--
Bruce Ide
flyingrhenqu...@gmail.com
you how to do that, you could probably do it with a beanshell function or
something.
--
Bruce Ide
flyingrhenqu...@gmail.com
chance for me to look into how to write predefined
functions in jmeter! :-D
--
Bruce Ide
flyingrhenqu...@gmail.com
I thought I saw that IterationListener had an iterationEnd event added at
some point but I'm not seeing it in the checkout of the 2.4 source I have.
Was that added in a later version?
--
Bruce Ide
flyingrhenqu...@gmail.com
.
--
Bruce Ide
flyingrhenqu...@gmail.com
x27;d start checking potential hardware issues,
because your files should not be getting corrupted like that. I'd start with
a virus checker and maybe run disk and RAM diagnostics.
--
Bruce Ide
flyingrhenqu...@gmail.com
ng on what the string in the variable is:
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#If_Controller
--
Bruce Ide
flyingrhenqu...@gmail.com
MX file?) open in wordpad. Is that
something you do often? Are you changing and saving it with wordpad? Perhaps
that's what's corrupting the XML.
--
Bruce Ide
flyingrhenqu...@gmail.com
t
could potentially cause that sort of problem.
If none of those apply, I'm stumped for possible jmeter causes. At that
point I'd start looking at the system to see if there are signs of a faulty
disk or other hardware that might be causing files to become corrupt.
--
Bruce Ide
flyingrhenqu...@gmail.com
I really get the feeling you're leaving out key pieces of information that
would actually enable us to solve this problem. What version of Jmeter are
you using? And what version of Jmeter was your test written with?
--
Bruce Ide
flyingrhenqu...@gmail.com
hough. You can also use the SampleResult in a BSF
sampler to write to either the results or to the sampler label.
--
Bruce Ide
flyingrhenqu...@gmail.com
Sweet, that may be all I need to control my variable explosion. I'm
definitely going to experiment with that!
--
Bruce Ide
flyingrhenqu...@gmail.com
ng of my own.
Does anyone know of a controller or other jmeter plugin that would provide
"local variable" functionality?
--
Bruce Ide
flyingrhenqu...@gmail.com
I am a scientist, sir! I can never be absolutely positive! ;-)
--
Bruce Ide
flyingrhenqu...@gmail.com
setting JAVA_HOME to
the results of the command "dirname `which java`". I'm almost completely 99%
certain that if java is in your path, jmeter will figure it out without this
step being needed.
--
Bruce Ide
flyingrhenqu...@gmail.com
whole thing into
memory at once. I ended up just making them give me a smaller file to work
with.
--
Bruce Ide
flyingrhenqu...@gmail.com
That would actually be pretty cool. I'm always forgetting to disable my
listeners. Most of my tests don't run out of memory anyway, but there's no
point in using it if you don't have to!
--
Bruce Ide
flyingrhenqu...@gmail.com
Yep. Each user is executed in a separate thread, and gets a different
session with the authorization manager. Just set number of users to 10 in
your thread group and you're done.
--
Bruce Ide
flyingrhenqu...@gmail.com
x27;re interested in...
--
Bruce Ide
flyingrhenqu...@gmail.com
esult (Hell you
probably could down-cast it to SampleResult when you're done with it, which
I would do just to make the OOP crowd burst a blood vessel!)
--
Bruce Ide
flyingrhenqu...@gmail.com
edirects".
The redirects now show up as sub-samples to the main sampler, which works
fine for my test.
--
Bruce Ide
flyingrhenqu...@gmail.com
t
jmeter's performance. It sounds like you're dealing with a fairly small
response, though.
I often find that looking at the response in the "view results tree" sheds
some light on the problem. Usually it's something my regex didn't anticipate
being possible.
--
Bruce Ide
flyingrhenqu...@gmail.com
Some people, when confronted with a problem, think "I know! I'll use regular
expressions!" Now they have two problems.
--
Bruce Ide
flyingrhenqu...@gmail.com
sn't a jmeter plugin that
handles ntlm already, someone could probably use the source as the
foundation for one. Assuming that there's not a suitable java library that
could be used instead...
--
Bruce Ide
flyingrhenqu...@gmail.com
ly need to?
Heh heh heh
--
Bruce Ide
flyingrhenqu...@gmail.com
they're "cool"?
* Do I really understand what threads are used for?
Note that if only the answer to #3 is incorrect, the "Don't do that...
unless you really want to" rule kicks in, but generally you're on your own
if you do it anyway! :-D
--
Bruce Ide
flyingrhenqu...@gmail.com
f which was you actually could only have
a single session. If you tried to have more than that, they'd overwrite each
other in unpredictable ways.
--
Bruce Ide
flyingrhenqu...@gmail.com
e anything with subsamples myself, I've just noted that there
seems to be that capability while digging around in the source code.
--
Bruce Ide
flyingrhenqu...@gmail.com
I usually make a class BeanInfo class with the bean information. I haven't
yet needed to make a more complex UI than just "Enter a string here". I
might have to soon, though!
--
Bruce Ide
flyingrhenqu...@gmail.com
asses earlier rather than later,
and log everything. Then you can look in your jmeter log and see how far
your class got, or if it get anywhere at all.
--
Bruce Ide
flyingrhenqu...@gmail.com
went all crazy
with the emacs and the maven and all those kind thing.
--
Bruce Ide
flyingrhenqu...@gmail.com
arts, that is impossible to explain
the workings of to anyone who doesn't understand why a car engine is
beautiful.
--
Bruce Ide
flyingrhenqu...@gmail.com
ng terribly familiar with
swing than because of jmeter.
--
Bruce Ide
flyingrhenqu...@gmail.com
pplication. You could even maintain
different datasets to simulate light or heavy usage.
--
Bruce Ide
flyingrhenqu...@gmail.com
s would be
more work than writing a webapp to limit access to the system.
Though now that I think about it, having 40 separate admin users running in
the system at the same time might uncover other problems, so perhaps I
should reconsider, after all!
--
Bruce Ide
flyingrhenqu...@gmail.com
o anything else while requesting tiles, or if you want it
to start requesting tiles at one point in the test, and then stop later on
(While still doing other stuff,) that could get more complicated.
--
Bruce Ide
flyingrhenqu...@gmail.com
solution would be to modify the application to
print an error that the user is already logged in. Jmeter could detect that
with a XPath or Regex postprocessor and retry until the system allows it to
log in. We'll see if the intersection between ideal and real world allows
for that!
--
Bruc
ge request, requesting that the
behavior of the app be modified to allow multiple administrative logins or
print an error message if the same admin user attempts to log in instead of
kicking the current session out. But then I'd have to talk to developers
(heh heh heh.)
--
Bruce Ide
flyingrhenqu...@gmail.com
ing looks promising. I'll have to throw
together a jmeter test to exercise it in the morning.
--
Bruce Ide
flyingrhenqu...@gmail.com
y specific config/settings to be made?
>
You just want it to do the same thing at the same time in all threads? You
could use a synchronizing timer to block all threads until they all reach
the timer.
http://jakarta.apache.org/jmeter/usermanual/component_reference.html#Synchronizing_T
sults did, in fact, retrieve columns from the database. Then
look at the debug sampler and see how it created the variables for you to
use.
--
Bruce Ide
flyingrhenqu...@gmail.com
manager
panel, or a "send expired cookies" checkbox to the httpclient. Probably
wouldn't be a huge amount of coding, and would probably be only vaguely
atrocious.
Or perhaps a sampler or postprocessor that allows you to manipulate explicit
cookie values? That'd be a bit more work, but might be more palatable.
--
Bruce Ide
flyingrhenqu...@gmail.com
7;s probably also
some code in httpclient that checks to see if they're expired before sending
them back.
I'd be really hesitant to change the behavior of the test environment to
mask a bug you uncovered, though. Sending expired cookies IS a bug, and it's
something the guys running the server should fix.
--
Bruce Ide
flyingrhenqu...@gmail.com
The only thing I can think of that would cause that is that your web browser
isn't using the proxy for local addresses. If you're using firefox, do you
have anything at all in the "No proxy for:" in your network settings? If so,
remove it and try again.
--
Bruce Ide
flyingrhenqu...@gmail.com
ing the assertion in place and supplying the correct login
and password. When you go to the next page, the assertion will look for the
error text, not find it and cause a test failure.
--
Bruce Ide
flyingrhenqu...@gmail.com
ryone
thinks they're so simple that examples aren't documented.
--
Bruce Ide
flyingrhenqu...@gmail.com
Add another response assertion?
--
Bruce Ide
flyingrhenqu...@gmail.com
ing the same one over and over again? If it's created by the
Javascript side of the application, I would expect the java side to just
accept whatever you tell it that ID is.
--
Bruce Ide
flyingrhenqu...@gmail.com
ient side UI and happily accept anything you send to
the backend if you bypass that checking...) but it does not do any sort of
UI validation. So you'll still need someone to go through the UI and make
sure buttons still work and stuff.
--
Bruce Ide
flyingrhenqu...@gmail.com
u said, jmeter variables are local to the thread, so
changes made in other threads will not be seen in your current one (Barring
trickery :-)
--
Bruce Ide
flyingrhenqu...@gmail.com
only a class or interface name changes.
--
Bruce Ide
flyingrhenqu...@gmail.com
1 - 100 of 145 matches
Mail list logo