Date: Wed, 17 Jun 2015 21:38:34 -0400 From: Mark Saad <nonesuch@longcount.org> To: Steven Hartland <steven@multiplay.co.uk> Cc: Mark Martinec <Mark.Martinec+freebsd@ijs.si>, "freebsd-fs@freebsd.org" <freebsd-fs@freebsd.org> Subject: Re: TRIM erases user data Message-ID: <CAMXt9NaBpi7VK2OUqOUNPXP3k3yoVy2YrVBO5ru-CsjZFEyYyw@mail.gmail.com> In-Reply-To: <99F249A6-F285-44D0-B003-5A7C96D20FD7@multiplay.co.uk> References: <CAD2Ti2-dcSvJvLuQgy=hn3wB8C8eho4Y3moxGP3YSW3AN=_Axg@mail.gmail.com> <49FEE0B5-D924-4FB6-889B-54F18667239B@multiplay.co.uk> <d12986a58be6f2df232d822eb12b53c3@mailbox.ijs.si> <7B981BB3-800B-4F59-B842-4D08ECDC5228@multiplay.co.uk> <B2127775-1BC2-4651-888A-8A9C4E367C1E@longcount.org> <99F249A6-F285-44D0-B003-5A7C96D20FD7@multiplay.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jun 17, 2015 at 8:30 PM, Steven Hartland <steven@multiplay.co.uk> wrote: > Just for the record illumos doesn't have any ZFS TRIM support. > > There are lots of things in illumos that applys to :(. In any case I wasn't aware that illumos zfs was sans-trim . > > On 17 Jun 2015, at 17:25, Mark Saad <nonesuch@longcount.org> wrote: > > > > > > > >> On Jun 17, 2015, at 7:55 PM, steven@multiplay.co.uk wrote: > >> > >> We've used Samsung 840 Pro's on FreeBSD with ZFS for a long time (which > has TRIM enabled by default) so as far unless this has been broken on a FW > or HW version we've not used then I have no reason to believe we're > effected by this issue at this time. > > > > I was using 840 pros for zfs on freebsd 9.3 , 10.1 for 2 years with no > issues . It was mostly for archival video storage and used as l2arc . > > > > I am working with omnios and smartos (illumos / opensolaris) almost > exclusively with zpools on ssds of various vendors and I haven't see this > issue or similar issues . > > > > I am still using 4 840ev on a 10.1 zfs box for the pool , and one for a > l2arc and it appears to be working as expected . > > > > For what my opinion is worth of this was some weird hardware bug , it > wouldn't just effect Ubuntu. I think Ubuntu or better said the culture of > Ubuntu might have a hand in this blog post . > > > >> > >> Sent from my iPad > >> > >> On 17 Jun 2015, at 15:58, Mark Martinec <Mark.Martinec+freebsd@ijs.si> > wrote: > >> > >>>>> On 16 Jun 2015, at 18:51, grarpamp <grarpamp@gmail.com> wrote: > >>>>> https://blog.algolia.com/when-solid-state-drives-are-not-that-solid/ > >>>>> > https://github.com/torvalds/linux/blob/e64f638483a21105c7ce330d543fa1f1c35b5bc7/drivers/ata/libata-core.c#L4109-L4286 > >>>>> > http://www.aerospike.com/docs/operations/plan/ssd/ssd_certification.html > >>> > >>> steven@multiplay.co.uk wrote: > >>>> This issue centers around queued TRIM requests at the ATA layer, which > >>>> is an extension that allows NCQ support for TRIM in the SATA 3.1 spec. > >>>> This is not something FreeBSD currently supports so is unaffected by > >>>> the issue at this time. > >>> > >>> Are you sure? The article explicitly states they were not using > >>> queued TRIM: > >>> > >>> | A lot of discussions started pointing out that the issue is related > >>> | to the newly introduced queued TRIM. This is not correct. The TRIM > >>> | on our drives is un-queued and the issue we have found is not related > >>> | to the latest changes in the Linux Kernel to disable this features. > >>> > >>> > >>> Mark > >>> _______________________________________________ > >>> freebsd-fs@freebsd.org mailing list > >>> http://lists.freebsd.org/mailman/listinfo/freebsd-fs > >>> To unsubscribe, send any mail to "freebsd-fs-unsubscribe@freebsd.org" > >> _______________________________________________ > >> freebsd-fs@freebsd.org mailing list > >> http://lists.freebsd.org/mailman/listinfo/freebsd-fs > >> To unsubscribe, send any mail to "freebsd-fs-unsubscribe@freebsd.org" > > > > > > --- > > Mark Saad | nonesuch@longcount.org > -- mark saad | nonesuch@longcount.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAMXt9NaBpi7VK2OUqOUNPXP3k3yoVy2YrVBO5ru-CsjZFEyYyw>