Date: Fri, 22 Apr 2016 23:59:36 -0600 From: Warner Losh <imp@bsdimp.com> To: Ivan Klymenko <fidaj@ukr.net> Cc: FreeBSD Current <freebsd-current@freebsd.org> Subject: Re: Heads up Message-ID: <CANCZdfqpiSMaeecf56bG22rXryXK3Gv-WK7eAmkzoWAV=HR9Fg@mail.gmail.com> In-Reply-To: <20160422213332.7e031073@nonamehost.local> References: <CANCZdfpnYnVrvhNagYUT9RhAuC1AMCrxh=GCt8RKT0bqxuJybw@mail.gmail.com> <20160422213332.7e031073@nonamehost.local>
index | next in thread | previous in thread | raw e-mail
On Fri, Apr 22, 2016 at 12:33 PM, Ivan Klymenko <fidaj@ukr.net> wrote: > On Thu, 14 Apr 2016 16:42:33 -0600 > Warner Losh <imp@bsdimp.com> wrote: > > > The CAM I/O scheduler has been committed to current. This work is > > described in > > https://people.freebsd.org/~imp/bsdcan2015/iosched-v3.pdf though the > > default scheduler doesn't change the default (old) behavior. > > > > One possible issue, however, is that it also enables NCQ Trims on ada > > SSDs. There are a few rogue drives that claim support for this > > feature, but actually implement data corrupt instead of queued trims. > > The list of known rogues is believed to be complete, but some caution > > is in order. > > > > Warner > > Hi. > > I have the small issue. > I have one hard drive with zfs. > If i beginning cloning the hard drive in VirtualBox machine - this > process is very slow 1,5-2 hours. > Without CAM_NETFLIX_IOSCHED options - this process two or three times > faster. > With this is possible to do something? > I don't know. I've not investigated this in enough detail to know. One thing you might change is the read bias from 300 to 1, which may be more appropriate for your workload. Warnerhome | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfqpiSMaeecf56bG22rXryXK3Gv-WK7eAmkzoWAV=HR9Fg>