News for hackers. Stuff that matters.
Re:AMD EPYC Performance Testing with systemd by dijit on Thursday July 12, 2018 @05:03PM (#17518590)
I'm not an expert, I'm a lowly systems admin-

But I imagine the difference is that size of systemd itself and how mysql does fork().

SystemD is 1.5MB itself on my systems where I have it, but upstart (for example) is 148KB on centos 6.

Since an AMD Epyc has roughly 64Mb of L3 Cache, a larger binary would not have to be evicted from L3 cache as often.

One of Intels generally powerful all-rounder CPUs (2687Wv4) only has 30Mb of "Smart-cache" (which is fancy speak for; not that much)

A complete guess on my part though..

Re:AMD EPYC Performance Testing with systemd by dsr_ on Thursday July 12, 2018 @06:13PM (#17519053)
I'm upvoting you because even though you turn out to be wrong, you appear to have been sincere about it. You shouldn't be punished, even in fake internet points, for offering a hypothesis in good faith.
Re:AMD EPYC Performance Testing with systemd by michaelmrose on Thursday July 12, 2018 @07:20PM (#17519423)
I don't think imaginary internet points serve entirely or even mostly to punish.

It served to highlight valuable content and in some venues to weed out low value content entirely. This is more significant in large bodies of comments. Example highlighting the most useful 20 comments out of 700.

Interestingly sometimes wrong information can lead to good threads where people provide helpful corrections. In this case the thread is valuable for stirring up discussion even if the information may be wrong.

Re:AMD EPYC Performance Testing with systemd by michaelmrose on Sunday July 15, 2018 @02:56PM (#17536458)
An additional point it may be sometimes be more worthwhile to downvote incorrect information only if it seems unlikely to bear useful fruit and after providing enlightening information to the commentor if possible.
Re:AMD EPYC Performance Testing with systemd by plantain on Thursday July 12, 2018 @05:06PM (#17518611)
L1/L2/L3 Cache doesn't cache whole binaries, just hot pages from it.
Re:AMD EPYC Performance Testing with systemd by lederhosen on Thursday July 12, 2018 @05:33PM (#17518805)
not hot pages, just cache lines
Re:AMD EPYC Performance Testing with systemd by evanelias on Friday July 13, 2018 @01:49AM (#17520779)
> how mysql does fork()

mysqld uses a single-process architecture; once running, it's not calling fork(). It creates a thread per connection, rather than forking a new process per connection like postgres.