| || |
What version of LoadRunner (LR) or PerformanceCenter (PC) are you using?
What is the protocol you are recording?
If .NET, which version of .NET does the app require?
If .NET, which version of .NET do you have installed?
If OracleNCA with Oracle Forms Server, please list the version of Oracle Forms Server.
If HTML - are you using HTML-Advanced with URLs?
If URL mode:
- Concurrent groups, or
- Without Concurrent groups?
Which LoadRunner/PerformanceCenter feature (FPs) or service packs are you using?
VuGen Recording - are you using Old or New Recording Engine?
List here the Licensed Vuser type/ quantity that you are/were using for this specific issue /problem.
Is your support/maintenance contract current and active?
What platform(s) and Operating Systems are being used for load generators and controllers?
Windows XP version 2002 Service Pack 2
If you have filed a service request with HP/Mercury, what have they told you at this point with respect to your issue?
What reasons can make a file download incomplete.
I can see the file getting downloaded in the log but its incomplete.
Say if am expecting a 2mb file am getting only 525 kb file.
am i missing something?
note: i have the step download timeout set to 1000 sec
-Server resources from where you are downloading
As i am not a pro in network adminstration i am looking for some helpful tips here.
in sitescope i can see only monitor for 'network bandwith'.
is there any other monitor i can use to prove this network problem?
how i can measure them?
i have access only to the appserver is it enough to measure the latency between my machine (which i used to run the controller) and the server?
what ways i can prove its the network issue?
any input will be helpful.
Can you download the file manually via the app without Vugen/LoadRunner?
Even when i do manually i see this prblem.
The winzip brings a message saying ' the downloaded file is not in the archive'. if you downloaded try to download again.'
This message appears only if the download is incomplete it seems..
If i download the second time manually ..it works.
Even in the controller runs, some are is passing (around 30)and some are failing (around 70) for download (due to incomplete download-i have a check for the file size in script).
If i try to download big files, then i see all failures in the controller run.
It seems you are encountering network and/or server issues, but it is possible it may be related to the demands of your tests.
During your test using the LR native monitor or SiteScope, check that throughput is not consistently flat-lining at some point above zero AND below advertised bandwidth for your network(s). That would indicate some possible throttling and/or clogged network.
Bear in mind, if different networks are traversed between your load generators and the server architecture-under-test, they may have different bandwidths.
I would suggest you involve a network admin and/or acquire and use Ethereal or WireShark to get a close up look at what might be going on.
Aside from what is going on when you see this manually...
If your test is consuming more resources than are available such that you are stressing either or both of network or server architecture, then the issue is one of the demands of your test.
Note that other network appliances may be contributing to the issue.
after little bit analysis of the logs i came to know that the logic i have put in the scrit is executing certain statement n number (based on web_reg_find text) of times in vugen.
The same statement is executing n+10 number of times in controller which is consuming more resource or making more calls.
Why this difference?
ramaveda, as Jake said, I'd suggest that you focus exclusively on manual downloading of the file right now. If you can't perform the download manually as you'd like, there's no way you'll be able to do this automated-ly (not a word, but you get what I mean).
Focusing on the manual aspects of your process will yield you a clean template on which to build your automated script. Right now, with the manual issues, you do not know if you've got an actual clean record, so you must get the issue fixed first.
I can do manually but i too see failures when do manually.
If i do manually it fails 8th time or 10th time.
But in the controller its failing from the beginning..
why is it so?
Does the controller doing something different than manual steps. if so what it will be?
ramaveda, I'm sorry to say, but it really is impossible to tell without sitting where you are sitting. For problems like this, the best we can do as volunteers to try to help you is suggest general rules of problem diagnosis.
I'd suggest that your best clue is that you know manual does fail, and you can reproduce its failure. Given that fact, if you can reproduce its failure and trace the issue there, you may get a lead on your load testing issue.
Though LR approximates production usage well, the load configuration is different from a "real" user in a number of ways. When you find the cause of the manual failure, you'll have a lead on what might be causing your automation failure. And you can think critically from there.
The specific cause of your issue can be any number of things - I'd highly suggest that you've got to start at the beginning. Your first step is to find the manual error - to do this, 1) Check any logs you may have on the manual process very closely, the answer may be in there 2) Use Wireshark to sniff the traffic and see what is happening at the network layer -or- if you're not comfortable w/ Wireshark, engage a network expert.