Welcome Serena Central users! CLICK HERE
The migration of the Serena Central community is currently underway. Be sure to read THIS MESSAGE to get your new login set up to access your account.
Frequent Contributor.. heldmacm Frequent Contributor..
Frequent Contributor..
163 views

LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Hello,

I am receiving some questions about the validity of the response time our LR HTML report shows vs. the amount of time it takes a user to do the same HTTP task. The LR report is coming from a single user run through the controller, and the the user is manually measuring the time to do the same in the application immediately afterwards. Not only is there a difference between these two, but the LR report is consistently showing 35-40 seconds for the transaction time while it takes 8-10 to perform manually. Since the user actions are within a transaction and there isn't think time in the transaction, I have no idea why this is happening.

The only thing I can think that could cause an issue is that when I recorded this HTTP script, the LR counter never completely stops even after the page is fully loaded in the browser (i.e. all on screen objects have loaded). Once the page appears fully loaded is when I stopped the transaction, so there shouldn't be much "extra" time factored in to the transaction. Can someone please give me a hand as to why there is such a discrepency?

Thank you.
0 Likes
8 Replies
Trusted Contributor.. jj_jena Trusted Contributor..
Trusted Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Multiple things could be analyzed here. If you check these options (there may be more) you may get the LR time to almost match actual time.
1) Have you disabled the Extra resources?
2) Have you checked Browser caching, simulate new user on each iteration from RTS?
3) Is it that for 1st iteration the time is higher and second iteration onwards the time is lesser?
4) Is login/logout kept in init and end actions?
5) Is extended log disabled?
0 Likes
Frequent Contributor.. heldmacm Frequent Contributor..
Frequent Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Thanks for the quick reply.  Please see the answers below:

1) Have you disabled the Extra resources? I have not - would you suggest I try doing so?
2) Have you checked Browser caching, simulate new user on each iteration from RTS? Simulate browser cache is checked and simulate new user is checked (but I'm only running the test for a single vuser through the controller, so there should only be one user)
3) Is it that for 1st iteration the time is higher and second iteration onwards the time is lesser? N/A as the transaction is in the Init section (see the answer to the fourth question for additional information)
4) Is login/logout kept in init and end actions? This specific business process is part of the login process and is in the init section.  It's measuring the time it takes a load a page after clicking the link to the page from the initial URL set in LR.  So when recording, the user navigates to an initial page, clicks a link (the transaction in question), and after the page with the clicked link loads, the Action section begins.
5) Is extended log disabled? I had extended logging enabled with only Parameter Substitution checked.  I reran the scenario with Standard Logging selected and did not observe any notable difference.

0 Likes
Trusted Contributor.. jj_jena Trusted Contributor..
Trusted Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Extra resources are all the static contents which are stored on the browser on client side. By not disabling it, you are forcing the static contents to be downloaded from server again which won't happen with real browser. So better to disable it.
0 Likes
Frequent Contributor.. heldmacm Frequent Contributor..
Frequent Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Thanks for the response.  I commented out all EXTRARES lines in the scripted (along with the files that follow it in the call), and I did not notice any difference running it this way through the controller with a single user.  Any other suggestions that would cause such a large difference between what LR is reporting when run as a single user and how long it actually takes a single user through a browser?

0 Likes
Trusted Contributor.. jj_jena Trusted Contributor..
Trusted Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

You may compare how many requests are being generated manually and through LR. You can use developer tool to check this manually. It might be that when you measure time manually, user is stopping the watch even though there are still responses (that can't be seen easily to the eye) coming from server.
0 Likes
Rick3 Contributor.
Contributor.

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Are both tests run on the same computer/agent?
0 Likes
Frequent Contributor.. heldmacm Frequent Contributor..
Frequent Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

Good question - I had been running the manual test locally, and the performance test is running on a generator.  I tried manually through the generator, and I'm seeing similar results as I am locally (roughly 8-10 seconds vs. 30+ seconds through LR).

0 Likes
Frequent Contributor.. heldmacm Frequent Contributor..
Frequent Contributor..

Re: LR 12.55 Transaction Time Significantly Longer than Manual Time for Single User

I suspect this is the problem, as Fiddler shows traffic still coming in after that point that I'd hit "stop" on the stopwatch.  If one or more of these requests are still running for some time after the point that the page has fully loaded in the browser, that would explain why we're seeing such a discrepancy in the response times between LR and doing the process manually.  Thanks for your help.

0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.