I’m more or less at same place here… the only difference is that after few hours of testing I was lucky to catch he issue once with Chrome and a little bit more info was revealed.
So that’s the issue: Still no pattern, but sometimes and some files are “stalled” during 10 to 20 seconds and overall performance is a crap.
What it’s wired is the issue appears much more in Firefox, than in Chrome.
And much more with 304 resources than 200 petitions.
And in the same server, it never happens to us with 2.3.6 branch.
And apache error log is clean.
Google define “stalls” as:
Stalled/Blocking: Time the request spent waiting before it could be sent.
This time is inclusive of any time spent in proxy negotiation. Additionally, this
time will include when the browser is waiting for an already established
connection to become available for re-use, obeying Chrome’s maximum six
TCP connection per origin rule.
I have few theories here:
a) KeepAlive issue? I disabled it and looks like the problem is still there.
b) Too much files & connections? Is there any recommended way to join CSS and JS in 2.x branch?
c) Network issue? Log is clean, but my be it’s related with iptables/bounding/networkcard issue?
Any feedback is welcomed.
Regards,
m.