core(config): include additional trace categories #12692
Conversation
Signed-off-by: Victor Porof <[email protected]>
victorporof
changed the title
| 'disabled-by-default-devtools.timeline.frame', | ||
| 'disabled-by-default-v8.cpu_profiler', | ||
| 'disabled-by-default-v8.cpu_profiler.hires', | ||
| 'latencyInfo', |
There was a problem hiding this comment.
how much larger do these typically make traces? Will we need to update things like minify-trace to strip new things (or include new things) in the near future?
There was a problem hiding this comment.
I would expect cpu_profiler.hires to be the largest of these by far on busy pages (which we were planning on enabling anyhow #8526). I think latencyInfo can be fairly chatty too if it's what I think it is.
Updating minify-trace to keep what we need from here, would be great! Could use that process to get some rough data on trace sizes for our common culprits (cnn.com, sfgate.com, theverge.com)
brendankenny
left a comment
brendankenny
left a comment
There was a problem hiding this comment.
LGTM, assuming no one else has an issue here.
A run on the verge gives a trace of
current Lighthouse - 50.1MiB
this branch - 60.7MiB
So, it's a significant jump (albeit from an already very large base :). Is that going to be an issue for anyone in terms of raw storage?
We'll definitely want to follow up with minify-trace pretty soon, I think.
|
Filed #12748 for updating minify-trace. The test failure is a known intermittent. |
Enough time for objections has probably passed :) we can just keep an eye out for any issues |
5 participants
Signed-off-by: Victor Porof [email protected]