Date: Tue, 13 Nov 2018 12:40:13 +0100 From: Borja Marcos <borjam@sarenet.es> To: Steven Hartland <killing@multiplay.co.uk> Cc: joe mcguckin <joe@via.net>, freebsd-fs@freebsd.org Subject: Re: ZFS flash systems Message-ID: <E7A19ECA-015A-43A2-907D-1A5923040C6E@sarenet.es> In-Reply-To: <CAHEMsqamBvTA=mTfkYxo-WdkF2_0D2RweAVAC%2BEMqOooTAziZw@mail.gmail.com> References: <B6E30DCC-2E05-4849-A04B-E8F740DF1370@via.net> <CAHEMsqamBvTA=mTfkYxo-WdkF2_0D2RweAVAC%2BEMqOooTAziZw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> On 11 Nov 2018, at 11:43, Steven Hartland <killing@multiplay.co.uk> = wrote: >=20 > It does perform well and it does have some specific knowledge when it = comes > to the underlying storage on FreeBSD specifically in knows rotating vs > non-rotating media. The worst issue is, unless I am wrong you won=E2=80=99t have hot swap. = At least when I tried the code for the =E2=80=9Cwarm swap=E2=80=9D ritual was not in place. You can=E2=80= =99t just unplug like you would a SAS disk. You need to put the device in quiescent mode before. Also, beware the NVMes and their TRIM implementation. Check firmware = versions and test in advance. I had a nasty surprise with Intel NVMEs some time ago. Borja.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E7A19ECA-015A-43A2-907D-1A5923040C6E>