Date: Fri, 31 Aug 2018 13:08:46 +1000 (AEST) From: Grant Gray <grant@gray.id.au> To: "M. Casper Lewis" <mclewis@genomecenter.ucdavis.edu> Cc: freebsd-fs <freebsd-fs@freebsd.org> Subject: Re: Failing ZFS log devices/panic Message-ID: <61612853.258505.1535684926024.JavaMail.zimbra@grantgray.id.au> In-Reply-To: <20180831030405.GZ1473@genomecenter.ucdavis.edu> References: <20180831003436.GW1473@genomecenter.ucdavis.edu> <707525919.257415.1535675999891.JavaMail.zimbra@grantgray.id.au> <20180831011607.GX1473@genomecenter.ucdavis.edu> <1121631966.257797.1535679312447.JavaMail.zimbra@grantgray.id.au> <20180831025805.GY1473@genomecenter.ucdavis.edu> <20180831030405.GZ1473@genomecenter.ucdavis.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
----- On 31 Aug, 2018, at 1:04 PM, M. Casper Lewis mclewis@genomecenter.ucdavis.edu wrote: > On Thu, Aug 30, 2018 at 07:58:05PM -0700, M. Casper Lewis wrote: >> On Fri, Aug 31, 2018 at 11:35:12AM +1000, Grant Gray wrote: >> > I've currently got an issue between some HGST SATA disks and a SAS3008 HBA >> > where mixing SAS and SATA disks on the same port results in intermittent >> > I/O errors on the HGST SATA disks, but not other SAS devices. >> >> This sounds like what we are seeing, but we're not mixing SAS and SATA. > > I forgot to add: Why are the log devices showing problems and not the > cache devices? Same drives, same bus. Different access pattern, sure. > > -- > M. Casper Lewis | mclewis@ucdavis.edu > Systems Administrator | Voice: (530) 754-7978 > Genome Center | > University of California, Davis | It could be a function of synchronous write load. Log devices must be coherent. Cache devices can back-fill (I don't know how the ZFS algorithm works, but I imagine pages will continue to hit the L1 ARC before and during demotion to the L2 ARC).
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?61612853.258505.1535684926024.JavaMail.zimbra>