Date: Fri, 18 Nov 2016 14:54:17 +0000 From: Matthew Seaman <matthew@FreeBSD.org> To: freebsd-questions@freebsd.org Subject: Re: testing SSD performance Message-ID: <36c4714c-f1ba-527e-b4e2-04072d1be9f9@freebsd.org> In-Reply-To: <9966b2a4-8812-3de1-8a9e-05bc96c8ef3c@kukulies.org> References: <9966b2a4-8812-3de1-8a9e-05bc96c8ef3c@kukulies.org>
next in thread | previous in thread | raw e-mail | index | archive | help
This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --qLsr3haj2ailg3vjUP54lO5ng3j4frFvw Content-Type: multipart/mixed; boundary="dxkV5v3LcCv8BjSjPqvbtkI7PKgWgcLdS"; protected-headers="v1" From: Matthew Seaman <matthew@freebsd.org> To: freebsd-questions@freebsd.org Message-ID: <36c4714c-f1ba-527e-b4e2-04072d1be9f9@freebsd.org> Subject: Re: testing SSD performance References: <9966b2a4-8812-3de1-8a9e-05bc96c8ef3c@kukulies.org> In-Reply-To: <9966b2a4-8812-3de1-8a9e-05bc96c8ef3c@kukulies.org> --dxkV5v3LcCv8BjSjPqvbtkI7PKgWgcLdS Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable On 11/18/16 14:37, Christoph P.U. Kukulies wrote: > I was running an older FreeBSD (9.2) for quite a while now with the roo= t > FS (and swap) an an SSD drive (100GB). >=20 > While upgrading the system to 11.0 and while observing quit elong > buildworld times I'm wondering myself if my SSD possibly might have > gotten degraded > some extent. I've read that SSD tend to run slow on writes over time, > especially when no provisions were taken to e.g. fill them up just up t= o > 2/3 of their max capacity to leave room for firmware storage management= =2E >=20 >=20 > Is there a quick tool to test overall performance (disk i/o and > processor overall system speed)? >=20 bonnie++ will give you some performance stats, although without a set of measurements on a new SSD to compare, its utility is likely to be limited= =2E The other utility that may be of use to you is smartmontools -- this will be able to tell you how much wear you've put on your device. You'll want to look for 'Media Wearout Indicator' or 'Wear Levelling Count' depending on the manufacturer. The other reason for long buildworlds is not having a great deal of RAM. C-compilation is both quite memory intensive in itself, and also does a lot of file-io, so having plenty of RAM available for filesystem caching will help a lot. Cheers, Matthew --dxkV5v3LcCv8BjSjPqvbtkI7PKgWgcLdS-- --qLsr3haj2ailg3vjUP54lO5ng3j4frFvw Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJYLxYaAAoJEABRPxDgqeTnjxwP/32mZG8Dgi3Am5G5lPSzLByh 3m+4jlU6LmhNqnKLvh0H7tjzfUvCztPOiiIer5HvBlwnTUOt4t5rJu/DOs8HD2KI XsX7HPGxGkosljNiSzQtQaOsf9eN6xz9Od761xc3QsmpmeDGTC2yFrMNE5Qm70cB pz+n+vfLB1HMxI0WvPlUFKwQ/QaYf1bs63qiTRBytSl0hZAvbRSsYfpLd7n7xptB yj8pYmQmdA37eMPCIxdCDg7sSti+iD3WYyswb1SWQ754PNmqKryZ1rXyLyGLzG+4 7a5MMT5EVpt5eR0ZgN1Pip4mYS73zHZx5z+1buVPwUpr1LWmWu709OS1BZbnMLx+ UQMAHI0YDTF9CUYDValwgrrTyv+YwWT99UPy63nBiUOVmrfOIE9P4i4vV/RT8sgh kiYD+/GUL5DN1VDgyiQsGl99QAcWtDTSmohYP27KtCDtpU07ZbFp3BULeSW9qT0l YmKRCHBsCcE6L+ozUbErwNNgjRj0NIT9QjybYhoSS8ws22q5D7ozz1wRmu6bD3PC IrBhMiU33iiGu0c3zNaQR4a3Lhlm2U8xktxwEBYmSdm5P4ekvTJiXiKPO9Q9KWbV 5wOLevk+tj27hMuov6padpdhhJAfmjesC0TtZtBm8R5jVtQETr+QO8ETRbPXO0Nw x2v0Uos4/s3dx+DZXiu2 =vQgS -----END PGP SIGNATURE----- --qLsr3haj2ailg3vjUP54lO5ng3j4frFvw--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?36c4714c-f1ba-527e-b4e2-04072d1be9f9>