Date: Wed, 25 Aug 2004 23:33:55 -0600 From: Scott Long <scottl@samsco.org> To: Poul-Henning Kamp <phk@phk.freebsd.dk> Cc: arch@freebsd.org Subject: Re: potential re change for 5.3? Message-ID: <412D7643.9000700@samsco.org> In-Reply-To: <51730.1093498385@critter.freebsd.dk> References: <51730.1093498385@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
Poul-Henning Kamp wrote: > In message <412D0868.9060203@freebsd.org>, Scott Long writes: > > >>I'm not suggesting anything different, just making a note of something >>that might be desirable in the future. In a way, I see GEOM as having >>the potential to be like Netgraph where it intercepts operations that it >>wants to process through it's framework and lets ones that it doesn't >>pass directly through without a decoupling through extra kernel threads. >>But that's only one possible strategy. Introducing the concept of a >>I/O scheduler that spawns KSE's to handle individual I/O requests is >>another possibility. > > > Well, the problem here is that requests which cannot be dealt with > due to resource shortages should be queued at the level where they > require least resources. A request queued inside a driver holds > far more resources than a request in the pure bio format at the > entrance to the driver for instance. Not necessarily. It's not hard for a driver to keep a request queued on the bioq and not consume and driver resources at all. > > And spawning KSE's in low memory situations is a recipe for suicide. > Probably true. Scott
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?412D7643.9000700>