Date: Fri, 05 Oct 2001 03:02:15 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: j mckitrick <jcm@FreeBSD-uk.eu.org> Cc: freebsd-advocacy@FreeBSD.ORG Subject: Re: sysadminmag update Message-ID: <3BBD8527.449C8537@mindspring.com> References: <20010720030757.A7504@dogma.freebsd-uk.eu.org> <3B585CC8.480BA14D@mindspring.com> <20010720174929.A17080@dogma.freebsd-uk.eu.org> <3B592F01.D9ED5404@mindspring.com> <20011004193802.E24976@dogma.freebsd-uk.eu.org>
next in thread | previous in thread | raw e-mail | index | archive | help
j mckitrick wrote: > > On Sat, Jul 21, 2001 at 12:28:01AM -0700, Terry Lambert wrote: > | j mckitrick wrote: > | > > | > here it is: > | > > | > http://www.sysadminmag.com/articles/2001/0108/0108q/0108q.htm > | > | The "tuning" they did for the sockets and other things > | is still wrong; most of those things are impossible to > | set at anything other than boot time. The numbers will > | go up, but the available number of objects sitting in > | the allocation pools will not. > > Is this documented? Several sysctl options say in the manpage that they > only work at boot time. It is not well documented. Most of the FreeBSD early code is not well documented, nor is its memory footprint, or what is used, when, and what's preallocated, where, and wehere it's managed. Matt Dillon's article in "Daemon News" on VM gave a small amount of information on one aspect of the issue (the least interesting aspect, to my mind, actually). I've gone through and documented all the early stage stuff for myself, and have been considering cleaning it up for a Dr. Dobbs Journal, Pentad Embedded Systems Journal, or a Daemon News technical article (listed in order of what they are willing to pay for articles on FreeBSD internals). It does _NOT_ include Bruce's kernel debugger stuff, which is itself strange, and detracts from the focus of the story. It would make two chapters in a FreeBSD Internals book, as well, if anyone is seriously interested in a collaboration on such a thing, and we are able to get a publisher lined up (be warned right now: it won't be a book on -current, if I'm involved: the target moves too fast, with interface changes and all, such that any internals book you write will be grossly out of date by the time it is published). -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3BBD8527.449C8537>