Date: Wed, 9 Jan 2013 17:47:32 -0800 From: Artem Belevich <art@freebsd.org> To: Borja Marcos <borjam@sarenet.es> Cc: FreeBSD Filesystems <freebsd-fs@freebsd.org>, John <jwd@freebsd.org> Subject: Re: rc.d script for memory based zfs intent log Message-ID: <CAFqOu6hQvpJ6Z=rWFwu1hp9Yq6pii_hrhNta=L0b81ChKfEtKA@mail.gmail.com> In-Reply-To: <AAEA3C3D-7FD5-4F26-A73B-47264F05E088@sarenet.es> References: <20130109023327.GA1888@FreeBSD.org> <AAEA3C3D-7FD5-4F26-A73B-47264F05E088@sarenet.es>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jan 9, 2013 at 7:15 AM, Borja Marcos <borjam@sarenet.es> wrote: > > In case of a crash, seems to be riskier than using sync=disabled on the datasets you need. What is the impact on the data integrity of a suddenly disappearing ZIL? Losing ZIL used to be fatal for the pool. I think in recent ZFS versions (v28?) you will only lose transactions that were not committed to the pool yet, but don't quote me on that.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFqOu6hQvpJ6Z=rWFwu1hp9Yq6pii_hrhNta=L0b81ChKfEtKA>