Failed to instrument the browser

WebPageTest Private Instances
1

Hello,
Whenever we are trying to run the test, we use to get the error "Failed to instrument the browser", instead of getting actual execution metrics/waterfall. This is happening every time whenever test is getting invoked on FF browser. Does the agent version get updated is issue or the issue with our OS type, we are using 64 bit OS and host is VM host
May i know the exact reason for this error.

Sometime i have also noticed that browser agent is taking very long time to launch the browser, please give the reason for this behavior as well

I am using my private WPT instance, and above behavior i have noticed specially in case of Firefox

2

Do tests using Chrome and IE work OK?

3

Thanks for quick replay, yeah chrome sometimes gives same error but on IE browser family its all working fine.
Error rate is huge on FF, please help since our work got hamper and effecting productivity

4

Any Update on this ? Still waiting for the resoluion

5

Is it a 32-bit version of Firefox (64-bit will not work)? What version of the test agent are you running (properties on wptdriver.exe)? The current version is 2.15.0.194. There haven’t been many changes recently but if you’re running an old enough version it could certainly explain it (here is info on how to update the agents: https://sites.google.com/a/webpagetest.org/docs/private-instances#TOC-Updating-Test-Agents )

6

pmeenan,
thanks for the help, the version of FF we are using is 32 bit only and recently we updated the wptdriver version from than we are not facing any such issues.
But though we started facing issue of TimeOutError, may i know the reason for this.
We tested huge bunch of URLs like 60 to 100, 100 to 150 and more than 200 bunch of urls at a time on single test. In these we noticed few results of few URLs will get fail and error on wpt results page we got was “TimeOutError” with no waterfall and no times (first byte, fully loaded etc times) .
Having long list of URLs in single test is the issue ?
Due to long server time and getting late response from the server making browser agent to stop and skip is the issue?

Please help

Regards
Syed

7

Depends on the version of the code you are running on the server. testStatus.inc used to time out tests if they didn’t complete within a period of time and a lot of different runs in the same test could trigger it. The latest version of the code just requires that any individual test run doesn’t stall for more than 10 minutes (I’m pretty sure that’s the logic anyway).

Breaking the test up into a batch of smaller tests will tend to work better for a bunch of reasons.