Thanks Shlomi. I believe this will help Support team.
↧
Re: AJAX TruClient IE v11 : Handle SSO login
↧
Regarding Virtual Table Server capabiltities
Hi,
If anybody can give me information on below , It would be great.
1) I am using VTS latest version which supports Load Runner 12.02 and need to know whether VTS server can handle concurrent virtual user access..Is it queueing user requests and then, serve? or whether it has capabilty to handle concurrent virtual users (say 50 users).
2) I use below code snippets to read column data and it worked. However, after few iterations, it was found that the data in the table was exhausted. Is there any VTS function onlyto read data so that data in the table is not exhausted?
int rc = 0;
lr_start_transaction("retrieve_column");
rc = lrvtc_retrieve_row();
lr_log_message("retrieve_row rc=%d\n", rc);
lr_log_message("%s", lr_eval_string("{CustomerID}"));
lr_end_transaction("retrieve_column", LR_AUTO);
↧
↧
Re: Data Center Migration- Network Virtualization Use
Hi,
"Can we use Network Virtualization to monitor the network packet loss or for latency?"
There are several methods, depending on the type of application (for example, consumer facing or internal, mobile or desktop), for discovering with which network conditions an application should be tested.
In some cases you can use information your company already has, latency/loss between different sites from the Networking team for example, and in others measuring the latency/loss (site to DC for example) with continues ping may suffice.
With an NV license, you can also use the NV Global Library in LR which provides mobile network conditions of different connections (e.g. 3G/4G) in many locations (i.e cities), and landline network conditions between some locations. Or the NV Network Capture server/agents which allow you to measure/monitor the latency/loss between locations (site to DC for example), and then export the results to a file which can be imported in LR.
Hope it helps, let us know if you have further questions.
Yaron
↧
Re: LoadRunner 12.53 / Sitescope 11.30/11.33 Integration
Exacty the same response.
Hello from Connect!
↧
Re: Regarding Virtual Table Server capabiltities
Hi
1.
50 concurrent users don't say much if you don't know how often the are requesting values. It will also depend on the sizing of the sever hosting the VTS, and network latency in between LGs and VTS.
In our environment we made som tests timing of a table with 30 000 row, and a script only timing and reading data from the VTS. With 3 vUsers we averaged in about 300 reads per second. So it is very lightwight
2.
lrvtc_retrieve_row()
This will consume the first row.
lrvtc_query_row(int rowIndex)
This will get the data from rowIndex in the same way as lrvtc_retrive_row, but will not consume it. Alternativly you have to push the data back with lrvtc_send_row1()
br /ola
Please mark post as solved if your problems or questions is/are resolved.
If this post was valuable to you, please consider kudo it.
↧
↧
Loadrunner VUGEN server Response is not coming properly.
Hi,
I am getting below server response whenever runs a perticular function. Due to that not able to proceed further.
Server replay response:
Xhttp://schemas.microsoft.com/net/2005/12/windowscommunicationfoundation/dispatcher/faultD��\+{+HM��:�)9��VV�V�V��s�:�SenderV�V� aRhttp://schemas.microsoft.com/net/2005/12/windowscommunicationfoundation/dispatcher�a�:�DeserializationFailedV�V�xmllang�en-US��The formatter threw an exception while trying to deserialize the message: There was an error while trying to deserialize parameter http://tempuri.org/:objLVProposalDataXml. The InnerException message was 'There was an error deserializing the object of type System.String. '�Y@objLVGeneralServiceAppData b8http://schemas.datacontract.org/2004/07/GeneralServiceBL i)http://ww' contains invalid UTF8 bytes.'. Please see InnerException for more details.
Thanks,
Abhijitk
↧
Re: LoadRunner Community Edition + TruClient error
Hi,
You are right, the 3rd party error was due to the NV component, which I have uninstalled.
I don't bother much about the 80% warning. The real problem is that I don't see any Page Hit or Connection, nothing... The test runs, creating the vusers, but they don't do anything at all...
Best,
Rui
↧
TruClient opening multiple browser windows.
Hey.
Running Truclient 12.53, patch 4 and using web protcol. This out of the blue started happening to me, when I log out of my applicaion while in develper mode, it pops up another 2 tabs. The tabs that open then kick the test out of standard mode. Also when I drag over the generic object step and try to ID and object its really slow and if I dont move my mouse at all it seems like more browser windows open every like 10 seconds. Any ideas on what can cause this or why it just stated to happen?
↧
Slow Performance when Running Load Generator on the same machine as the Controller
Hello,
I recently upgraded from LoadRunner 12.02 to 12.53 Patch #4. The hardware setup has been the same for years. It's a small environment with only 250 Web HTML virtual users. Two servers are used, each with the full install. One has the license installed and acts as the Controller. When test are run, both machines are used as Load Generators (50/50). Since upgrading, the response times are about twice as high and throughput is cut in half, but only for work done on the Load Generator that is also the Controller. This is even with a test using only 1 virtual user. If we swap machines and make the other act as the Controller, then that is the slower of the two Load Generators. Is there some configuration or setup that I can do to make performance just as good when the Controller and Load Generator are running from the same machine? Is it a new requirement that the Load Generators be seperate from the Controller?
Thanks,
Greg
↧
↧
Re: Slow Performance when Running Load Generator on the same machine as the Controller
Hello,
Its best pratice to seperate the load generator and controller for test execution. When running the test make sure you have enough resources on the LG machines.
Hope this helps you.
↧
Re: Data Center Migration- Network Virtualization Use
Thank you the suggestion.
In controller on NV, i have configured Virtual location as Newyrok to California and added the custom setting like Bandwidth and Latency.
Do i need to configure any other settings before running the test? Just adding the Virtual location can simulate the behavior as expecting(like client in Newyork and Severs in calif)?
Please advise
↧
Load Generator vs Controller at different network location
Hello Experts,
1)Is it advisble to have Controller machines and Load generator machines are different geo graphical locations and under different network?
2)Our Client has LG's moved to calif and users are in NY. At this moment our client is not willing to purchase any additional licenses for NV or other tool. So in this case i have asked our client not to move the LG machines to Calif and keep them at NY to simulate real user behavior. Is this this correct approach?
↧
Stress Testing with Truculent protocol & web protocol for 500 users
Hi,
We are stress testing the qliksense based application.we are able to record few navigation with web protocol in 1 script and able to record 5 flows with Truculent due to several Ajax call an client side activity. The application is developed I. Qliksense technology,angular is, etc., When we hit 500 users for truculent protocol having 5 script and 500 users for web protocol having 1 script, we got high response time of more than 130 sec after 50 minutes of ramp up...mean 1user every 3 sec......after 54 minutes,the application server CPU ranges between 70-90% and CPU of 40 %.....then application had so many errors,high response time and CPU down to 4%,memory stagnant at 40% , throughout and hits was down to very low after 54 min. Also we checked the application manually during test and the system shown crashed message...can we consider the 54th min as breakpoint of this application.
Please suggest.
Note:the memory would not get released until the next server restart happens.
Regards
Raghul
We are stress testing the qliksense based application.we are able to record few navigation with web protocol in 1 script and able to record 5 flows with Truculent due to several Ajax call an client side activity. The application is developed I. Qliksense technology,angular is, etc., When we hit 500 users for truculent protocol having 5 script and 500 users for web protocol having 1 script, we got high response time of more than 130 sec after 50 minutes of ramp up...mean 1user every 3 sec......after 54 minutes,the application server CPU ranges between 70-90% and CPU of 40 %.....then application had so many errors,high response time and CPU down to 4%,memory stagnant at 40% , throughout and hits was down to very low after 54 min. Also we checked the application manually during test and the system shown crashed message...can we consider the 54th min as breakpoint of this application.
Please suggest.
Note:the memory would not get released until the next server restart happens.
Regards
Raghul
↧
↧
Re: Loadrunner Error(29997): Cannot kill process or thread with moniker "1".
I have got a similiar problem. LR 12.53 and Patch 4 installed. When NV Insights ist activated in Controller, after the testrun is finished, nothing will happen for about 10 minutes. Then Analysis is called, and Controller shows up the abovementioned error "Cannot kill process or thread with moniker "1"."
↧
Re: Stress Testing with Truculent protocol & web protocol for 500 users
HI Raghul,
From the description below it sounds like the applicaiton under test does not sustain the workload.
Before concluding that this is indeed an application issue make sure that the Load Generators are not a bottleneck by checking the resource utilization on them as well. Also, try to perform the same steps the script is doing manually from a different machine which is not driving load to experience the response time on your own.
If the Load Generator is not a bottleneck and response time is slow also when you navigate the application manually you need to either reduce the workload or request the development team to improve performance.
If the Load Generator is the bottleneck you need to use more Load Generators
Regards,
Shlomi
↧
Re: TruClient opening multiple browser windows.
Hi,
Which browser are you using?
Can you check its' version by looking at the executable properties tab from the file system? firefox.exe is located under <LR_HOME>\bin\firefox and chrome.exe is located under <LR_HOME>\bin\chrome
Thanks,
Shlomi
↧
Re: Data Center Migration- Network Virtualization Use
Hi
Yes - that is all you need to do.
Good luck with your test :-)
Lior
↧
↧
Re: Slow Performance when Running Load Generator on the same machine as the Controller
Hi
First of all - the HW reccomendation for a full install has changed from 12.02 to 12.53, Please read the readme file of 12.53 in this location and see if you need to upgrade your machine HW.
2nd - make sure you are running on the correct operating system (again, this information is in the readme file).
once this is verified, see if that helped your workload to improve.
Good Luck,
Lior
↧
Re: Load Generator vs Controller at different network location
Hi Nik,
Please see my reply inline
Thanks,
Shlomi
nik_0315000 wrote:Hello Experts,
1)Is it advisble to have Controller machines and Load generator machines are different geo graphical locations and under different network? [Shlomi] LoadRunner supports locating Load Generators in different geo locations. While it is better locating them in the same location it is not supposed to be a showstopper. The customer just need to make sure the network does not have too long delays and can sustain the results transfer across the locations.
2)Our Client has LG's moved to calif and users are in NY. At this moment our client is not willing to purchase any additional licenses for NV or other tool. So in this case i have asked our client not to move the LG machines to Calif and keep them at NY to simulate real user behavior. Is this this correct approach? [Shlomi] the best practice is to simulate load from the same location the real users are going to come from.
↧
Issue with IE11 - True Client Protocol : multiple browser windows
Hi All,
I am new to LoadRunner and True Client Protocol. In our application we have a scenario like.. After clicking some button a new window opens have to capture the time taken to load that new window..here in IE11 with true client protocol new empty tab is opening while recording and reply once I click the button. so here I am unable to check whether the required window is loaded or not,
it works for chrome .. seems some settings issue for IE11.
Could you please help me on this.
Thanks,
Khaja Shaik.
I am new to LoadRunner and True Client Protocol. In our application we have a scenario like.. After clicking some button a new window opens have to capture the time taken to load that new window..here in IE11 with true client protocol new empty tab is opening while recording and reply once I click the button. so here I am unable to check whether the required window is loaded or not,
it works for chrome .. seems some settings issue for IE11.
Could you please help me on this.
Thanks,
Khaja Shaik.
↧