We’ve got a significant (<1 second) gap in our waterfall in the IE8 test from Los Angeles, and we’re not sure where the issue is. I’ve read what I can find on the forum, and it seems this usually is caused by high CPU Utilization caused by bad JS.
But the CPU Utilization on tests on two of our key pages (below) show that CPU usage stops (literally zero), then there’s a full second of nothing, then the page seems to hit ‘document complete’, then the post-loading components start loading.
Also, if anything else strikes your eye that we could improve, I’d certainly appreciate it (and we already know we need to get a CDN going)!
Thanks in advance for any thoughts/recommendations you might have!
Do you only see it from Los Angeles? Looks like it shows up both in the first and repeat views so it looks pretty reproducible (may want to do more than 1 run to be sure though).
Given that it’s between the clicktale js loading and the clicktale gif firing I’d look there first but since it is happening at the end right before onload it could be from just about anything.
Do you see it in Chrome as well? If so then recording a dev tools timeline trace can help tell you what is going on. If it is only in IE and you can get it to happen from Dulles then you can use the Dynatrace-instrumented browser and capture a full trace of what IE is doing.