Date: Mon, 5 Jul 2021 16:29:00 +0100 From: Pete French <petefrench@ingresso.co.uk> To: Alan Somers <asomers@freebsd.org> Cc: Stefan Esser <se@freebsd.org>, FreeBSD Stable Mailing List <freebsd-stable@freebsd.org> Subject: Re: ZFS + mysql appears to be killing my SSD's Message-ID: <bfee43a2-87e3-ab4a-abfb-7ebe38f2d810@ingresso.co.uk> In-Reply-To: <CAOtMX2gWnEEKHHaDf0GmfC1wKAUqHK3ptNUxm23UFBsD_KHwKg@mail.gmail.com> References: <89c37c3e-22e8-006e-5826-33bd7db7739e@ingresso.co.uk> <2fd9b7e4-dc75-fedc-28d7-b98191167e6b@freebsd.org> <9c71d627-55b8-2464-6cc9-489e4ce98049@ingresso.co.uk> <CAOtMX2gWnEEKHHaDf0GmfC1wKAUqHK3ptNUxm23UFBsD_KHwKg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05/07/2021 15:51, Alan Somers wrote: > If you're using 4K sectors anyway, then GELI does not create any extra > write amplification. But if you're extra paranoid and you use the "-T" > option to "geli attach", then GELI will block TRIM commands. That could > hurt SSD lifetime. But I don't think "-T" is the default. You are > using 4K sectors, right? ZFS's ashift is set to 12? > -Alan Yup, 4k sectors, and I initialised geli telling it to present as 4k sectors. Am glad to hear that it wont amplify writes - thats what I assumed, but nice to have it confirmed. -pete.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bfee43a2-87e3-ab4a-abfb-7ebe38f2d810>