Some websites makes the browser stop working on the agent, and the cpu uses 100% on all the cores. It hapens on Chrome, Firefox, IE9 and on windows server 2008 rc2 x64, win7 x64 and win7 x86. The only thing to do is to wait for the wpt driver to kill the job, and the result will be empty. It is not every time it happes, I can run several test on the same site before the problem occurs.
It is especially on content heavy sites fx. the danish news paper site “eb.dk” or “cnn.com”.
I haven’t seen the issue. Can you tell if it is at the end of the test? That is when the agent code does the optimization checks and it re-compresses every image and gzips every response to see if there are any possible savings. If you are running on a really low-end system and testing a very content-rich site I can see that taking quite a long time (usually it takes just a few seconds, even for a site like cnn).
Yes, it happens at the end of the test. It looks like it is waiting for some content on the page, but I’m not sure it is the case. The system runs on a dual xeon cpu setup in vmware, so guess there is enough power.
This looks like the same issue as i have “Issues with wptdriver and complex pages”
Only happens on content heavy sites. Urlblast works perfectly but anything using wptdriver crashes 8 out of 10 times on content-heavy sites (ours is www.vg.no). Light sites has no issues.