Date: Sun, 4 Jun 2017 17:18:29 +0300 From: Slawa Olhovchenkov <slw@zxy.spb.ru> To: Warner Losh <imp@bsdimp.com> Cc: Allan Jude <allanjude@freebsd.org>, FreeBSD Current <freebsd-current@freebsd.org> Subject: Re: Time to increase MAXPHYS? Message-ID: <20170604141829.GD3182@zxy.spb.ru> In-Reply-To: <CANCZdfpD3G8gR=C2_AekM6VeJ6dzKOnP820OOoF1M_eS0MfJ3g@mail.gmail.com> References: <0100015c6fc1167c-6e139920-60d9-4ce3-9f59-15520276aebb-000000@email.amazonses.com> <972dbd34-b5b3-c363-721e-c6e48806e2cd@elischer.org> <3719c729-9434-3121-cf52-393a4453d0b2@freebsd.org> <CANCZdfrkc1ERKnJr4JzHpePmU%2BrN5JOgAVePCShPHLDCAE19=w@mail.gmail.com> <CANCZdfpD3G8gR=C2_AekM6VeJ6dzKOnP820OOoF1M_eS0MfJ3g@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jun 03, 2017 at 11:49:01PM -0600, Warner Losh wrote: > > Netflix runs MAXPHYS of 8MB. There's issues with something this big, to be > > sure, especially on memory limited systems. Lots of hardware can't do this > > big an I/O, and some drivers can't cope, even if the underlying hardware > > can. Since we don't use such drivers at work, I don't have a list handy > > (though I think the SG list for NVMe limits it to 1MB). 128k is totally > > reasonable bump by default, but I think going larger by default should be > > approached with some caution given the overhead that adds to struct buf. > > Having it be a run-time tunable would be great. > > > > Of course 128k is reasonable, it's the current default :). I'd mean to say > that doubling would have a limited impact. 1MB might be a good default, but > it might be too big for smaller systems (nothing says it has to be a MI > constant, though). It would be a perfectly fine default if it were a > tunable. Some cloud providers limit IOPs per VM, for this cases MAXPHYS must be large as posible.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20170604141829.GD3182>