Date: Fri, 18 Jul 2008 14:12:08 +0200 From: Alexander Leidinger <Alexander@Leidinger.net> To: Peter Jeremy <peterjeremy@optushome.com.au> Cc: Doug Barton <dougb@freebsd.org>, David Naylor <naylor.b.david@gmail.com>, freebsd-current@freebsd.org Subject: Re: rc improvements (wanted?) Message-ID: <20080718141208.21091i4jkh44jc74@webmail.leidinger.net> In-Reply-To: <20080718071806.GV62764@server.vk2pj.dyndns.org> References: <200807172056.08835.naylor.b.david@gmail.com> <487FCA89.2010308@FreeBSD.org> <20080718083725.97823be0tg13fn6s@webmail.leidinger.net> <20080718071806.GV62764@server.vk2pj.dyndns.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Quoting Peter Jeremy <peterjeremy@optushome.com.au> (from Fri, 18 Jul =20 2008 17:18:07 +1000): > On 2008-Jul-18 08:37:25 +0200, Alexander Leidinger =20 > <Alexander@Leidinger.net> wrote: >> Are you aware that the parallel starting in Solaris 10 reduced the >> booting time by a nice percentage? > > Given that Solaris boots in geologic time, this probably wouldn't > be difficult. How do you define "booting Solaris"? Do you include the extensive =20 tests prior to loading the kernel into this? I'm not talking about the =20 time a 25k needs (even when you reducing the amount of testing on the =20 system controller, it takes a long while until it reaches a state =20 which I would call the start of the boot of the OS). We are talking =20 about the pure userland part of booting. What is done during the =20 startup of important programs in Solaris is not unreasonable (and =20 similar/comparable between Solaris versions), and still, there's a =20 nice difference between Solaris 9 and 10 if you count the time until =20 you can start to do useful stuff. >> If yes, do you expect that FreeBSD >> behaves significantly different or do you "just" want to see numbers? > > Parallel starting is not guaranteed to be an improvement. Starting a > whole pile of processes that are I/O bound during initialisation > (think squid or some databases) may be worse than starting them one > at a time. Likewise, a whole pile of processes that are CPU bound It depends, think about independent disks and or keeping the squid =20 data in RAM (e.g. tmpfs). But this doesn't matter, we will always be able to come up with =20 situations where the parallel start is not a good idea. We don't come =20 by default with such a situation and I'm sure a lot of configs out =20 there that don't fall into this class. Based upon your argument we =20 could say we can not enable parallel starting even if we see it is an =20 improvement for the reboot after the installation. What I wanted to know is if there's an substantial argument (it can =20 not behave similar to Solaris, because of A and B), or if he "just" =20 wants to know what the difference on FreeBSD is. > will just thrash the scheduler. (Though parallel starting of I/O and > CPU bound processes should be a win). You forgot about round-trip-time bound processes (basically processes =20 which wait for an event to occur before they say they are successfully =20 started), and we have several of them. Bye, Alexander. --=20 Those who hate and fight must stop themselves -- otherwise it is not stopped. =09=09-- Spock, "Day of the Dove", stardate unknown http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID =3D B0063FE7 http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID =3D 72077137
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080718141208.21091i4jkh44jc74>