opensuse
[Arriba] [Todas las Listas]

Re: [opensuse] Mundo real Sata-disco de II-a-disco throughput? 77MB/sec

To: opensuse@xxxxxxxxxxxx
Subject: Re: [opensuse] Mundo real Sata-disco de II-a-disco throughput? 77MB/sec es que sobre correcto?
From: "Brian K. White" <brian@xxxxxxxxx>
Date: Mon, 26 Sep 2011 19:46:47 -0400
Delivered-to: opensuse@xxxxxxxxxxxxxxxxxxx
Delivery-date: Mon, 26 Sep 2011 19:47:38 -0400
Envelope-to: traductor@xxxxxxxxxxx
In-reply-to: <4E7F94BC.5050602@suddenlinkmail.com>
List-archive: <http://lists.opensuse.org/opensuse/>
List-help: <mailto:opensuse+help@opensuse.org>
List-owner: <mailto:opensuse+owner@opensuse.org>
List-post: <mailto:opensuse@opensuse.org>
List-subscribe: <mailto:opensuse+subscribe@opensuse.org>
List-unsubscribe: <mailto:opensuse+unsubscribe@opensuse.org>
Mailing-list: contact opensuse+help@xxxxxxxxxxxx; run by mlmmj
Organization: Aljex Software Inc.
References: <4E7F94BC.5050602@suddenlinkmail.com>
User-agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:6.0.2) Gecko/20110902 Thunderbird/6.0.2
En 9/25/2011 4:53 PM, David *C. *Rankin Escribió:
*Guys,

Copiando vídeo entre dos *Seagate ST31500341CUANDO discos con *ext3
*filesystems consigo sobre 77.8 MB/*sec. Dado son supuestos para ser 300
MB/*sec:

"Es el 77 MB/*sec cierra al mundo real *max tendría que esperar?"

Entiendo que el 300 MB/*sec es el 'teórico *max' *throughput,
pero he nunca realmente pensamiento sobre lo que el 'real-mundial *max' tendría que
ser. Sé hay varios factores de hardware para numerar y medida de archivos
que impacto el índice, pero justo estoy probando para conseguir una idea si esto
es razonable.

La transferencia fue dirigida utilizando sobre 13 GB (sobre 1 *hr. De vídeo) de dato
en sobre 25 archivos:

15:37 *archangel:/*dat_*e/*dv> *du -*hcs *sony/*dcrv-1999.0[45].[023][167890]_*
1001*M *sony/*dcrv-1999.04.26_08-29-00.*dv
1.5*M *sony/*dcrv-1999.04.26_09-46-11.*dv
230*M *sony/*dcrv-1999.04.26_11-24-21.*dv
163*M *sony/*dcrv-1999.04.26_12-47-59.*dv
1001*M *sony/*dcrv-1999.04.26_15-41-40.*dv
1001*M *sony/*dcrv-1999.04.26_15-58-20.*dv
362*M *sony/*dcrv-1999.04.26_16-33-49.*dv
404*M *sony/*dcrv-1999.04.26_17-50-38.*dv
335*M *sony/*dcrv-1999.04.27_02-37-23.*dv
331*M *sony/*dcrv-1999.04.27_09-17-43.*dv
838*M *sony/*dcrv-1999.04.27_11-15-13.*dv
264*M *sony/*dcrv-1999.04.27_14-07-24.*dv
229*M *sony/*dcrv-1999.04.27_20-10-51.*dv
140*M *sony/*dcrv-1999.04.28_07-36-11.*dv
462*M *sony/*dcrv-1999.04.28_08-41-43.*dv
1001*M *sony/*dcrv-1999.04.28_10-52-18.*dv
584*M *sony/*dcrv-1999.04.28_11-07-26.*dv
373*M *sony/*dcrv-1999.04.28_14-50-35.*dv
1001*M *sony/*dcrv-1999.04.28_19-47-44.*dv
159*M *sony/*dcrv-1999.04.28_20-07-08.*dv
206*M *sony/*dcrv-1999.04.29_14-38-06.*dv
1001*M *sony/*dcrv-1999.04.29_20-56-26.*dv
431*M *sony/*dcrv-1999.04.29_21-18-38.*dv
258*M *sony/*dcrv-1999.04.30_22-33-44.*dv
546*M *sony/*dcrv-1999.05.01_17-14-07.*dv
13*G total

El cronometrando de la transferencia era:

15:37 *archangel:/*dat_*e/*dv> tiempo *cp -*uav /*dat_*e/*dv#unknown{^*sony,cinta-*idx.*txt}
/*dat_*f/*dv/
`/*dat_*e/*dv/*sony/*dcrv-1999.04.29_21-18-38.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.29_21-18-38.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_14-50-35.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_14-50-35.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.29_14-38-06.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.29_14-38-06.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_10-52-18.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_10-52-18.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_15-41-40.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_15-41-40.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_20-07-08.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_20-07-08.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_07-36-11.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_07-36-11.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.27_11-15-13.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.27_11-15-13.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_15-58-20.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_15-58-20.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_12-47-59.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_12-47-59.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.27_09-17-43.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.27_09-17-43.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.29_20-56-26.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.29_20-56-26.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.05.01_17-14-07.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.05.01_17-14-07.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_19-47-44.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_19-47-44.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_11-07-26.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_11-07-26.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.27_20-10-51.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.27_20-10-51.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_17-50-38.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_17-50-38.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.28_08-41-43.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.28_08-41-43.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_11-24-21.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_11-24-21.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_16-33-49.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_16-33-49.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.27_02-37-23.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.27_02-37-23.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.27_14-07-24.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.27_14-07-24.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_09-46-11.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_09-46-11.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.26_08-29-00.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.26_08-29-00.*dv'
`/*dat_*e/*dv/*sony/*dcrv-1999.04.30_22-33-44.*dv' ->
`/*dat_*f/*dv/*sony/*dcrv-1999.04.30_22-33-44.*dv'

Real 2*m47.980*s
usuario 0*m0.083*s
*sys 0*m27.982*s

Qué decir los expertos - razonable *throughput?


3*Gb es justo la interfaz, sólo el *cache en un paseo solo incluso se puede acercar que velocidad. El disco él incluso bajo las condiciones teóricas ideales de secuencial sostuvo leído del borde exterior del *platters en un fin alto 15*krpm disco y haciendo un crudo lee no pasando por cualquier *filesystem elevado, sólo puede lograr sobre 90MB.

77 leyendo archivos (no dato crudo) de un *filesystem y escribiendo a archivos en otro *filesystem, no podría ser el absoluto más alto estos paseos podrían lograr, podría haber algunas áreas de vuestro sistema total que podría ser *tweaked un poco, pero es bastante bien.

Para ir más que un minúsculo mordió más rápido necesitas *ssd o redada.

--
*bkw
--
A *unsubscribe, *e-correo: *opensuse+unsubscribe@xxxxxxxxxxxx
Puesto que órdenes adicionales, *e-correo: *opensuse+help@xxxxxxxxxxxx


On 9/25/2011 4:53 PM, David C. Rankin wrote:
Guys,

Copying video between two Seagate ST31500341AS disks with ext3
filesystems I get about 77.8 MB/sec. Given they are supposed to be 300
MB/sec:

"Is the 77 MB/sec close to the real world max I should expect?"

I understand that the 300 MB/sec is the 'theoretical max' throughput,
but I have never really thought about what the 'real-world max' should
be. I know there are various factors from hardware to number and size of
files that impact the rate, but I'm just trying to get an idea if this
is reasonable.

The transfer was conducted using about 13 GB (about 1 hr. of video) of
data in about 25 files:

15:37 archangel:/dat_e/dv> du -hcs sony/dcrv-1999.0[45].[023][167890]_*
1001M sony/dcrv-1999.04.26_08-29-00.dv
1.5M sony/dcrv-1999.04.26_09-46-11.dv
230M sony/dcrv-1999.04.26_11-24-21.dv
163M sony/dcrv-1999.04.26_12-47-59.dv
1001M sony/dcrv-1999.04.26_15-41-40.dv
1001M sony/dcrv-1999.04.26_15-58-20.dv
362M sony/dcrv-1999.04.26_16-33-49.dv
404M sony/dcrv-1999.04.26_17-50-38.dv
335M sony/dcrv-1999.04.27_02-37-23.dv
331M sony/dcrv-1999.04.27_09-17-43.dv
838M sony/dcrv-1999.04.27_11-15-13.dv
264M sony/dcrv-1999.04.27_14-07-24.dv
229M sony/dcrv-1999.04.27_20-10-51.dv
140M sony/dcrv-1999.04.28_07-36-11.dv
462M sony/dcrv-1999.04.28_08-41-43.dv
1001M sony/dcrv-1999.04.28_10-52-18.dv
584M sony/dcrv-1999.04.28_11-07-26.dv
373M sony/dcrv-1999.04.28_14-50-35.dv
1001M sony/dcrv-1999.04.28_19-47-44.dv
159M sony/dcrv-1999.04.28_20-07-08.dv
206M sony/dcrv-1999.04.29_14-38-06.dv
1001M sony/dcrv-1999.04.29_20-56-26.dv
431M sony/dcrv-1999.04.29_21-18-38.dv
258M sony/dcrv-1999.04.30_22-33-44.dv
546M sony/dcrv-1999.05.01_17-14-07.dv
13G total

The timing of the transfer was:

15:37 archangel:/dat_e/dv> time cp -uav /dat_e/dv/{sony,tape-idx.txt}
/dat_f/dv/
`/dat_e/dv/sony/dcrv-1999.04.29_21-18-38.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.29_21-18-38.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_14-50-35.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_14-50-35.dv'
`/dat_e/dv/sony/dcrv-1999.04.29_14-38-06.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.29_14-38-06.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_10-52-18.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_10-52-18.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_15-41-40.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_15-41-40.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_20-07-08.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_20-07-08.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_07-36-11.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_07-36-11.dv'
`/dat_e/dv/sony/dcrv-1999.04.27_11-15-13.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.27_11-15-13.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_15-58-20.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_15-58-20.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_12-47-59.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_12-47-59.dv'
`/dat_e/dv/sony/dcrv-1999.04.27_09-17-43.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.27_09-17-43.dv'
`/dat_e/dv/sony/dcrv-1999.04.29_20-56-26.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.29_20-56-26.dv'
`/dat_e/dv/sony/dcrv-1999.05.01_17-14-07.dv' ->
`/dat_f/dv/sony/dcrv-1999.05.01_17-14-07.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_19-47-44.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_19-47-44.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_11-07-26.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_11-07-26.dv'
`/dat_e/dv/sony/dcrv-1999.04.27_20-10-51.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.27_20-10-51.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_17-50-38.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_17-50-38.dv'
`/dat_e/dv/sony/dcrv-1999.04.28_08-41-43.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.28_08-41-43.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_11-24-21.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_11-24-21.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_16-33-49.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_16-33-49.dv'
`/dat_e/dv/sony/dcrv-1999.04.27_02-37-23.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.27_02-37-23.dv'
`/dat_e/dv/sony/dcrv-1999.04.27_14-07-24.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.27_14-07-24.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_09-46-11.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_09-46-11.dv'
`/dat_e/dv/sony/dcrv-1999.04.26_08-29-00.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.26_08-29-00.dv'
`/dat_e/dv/sony/dcrv-1999.04.30_22-33-44.dv' ->
`/dat_f/dv/sony/dcrv-1999.04.30_22-33-44.dv'

real 2m47.980s
user 0m0.083s
sys 0m27.982s

What say the experts - reasonable throughput?


3Gb is just the interface, only the cache on a single drive can even approach that speed. The disk itself even under the ideal theoretical conditions of sequential sustained read from the outer edge of the platters on a high end 15krpm disk and doing a raw read not going through any filesystem overhead, can only reach about 90MB.

77 reading files (not raw data) from one filesystem and writing to files on another filesystem, might not be the absolute highest these drives could attain, there might be some areas of your total system that could be tweaked a little, but it is quite good.

To go more than a tiny bit faster you need ssd's or raid.

--
bkw
--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx


<Anterior por Tema] Tema Actual [Siguiente por Tema>