Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.os.linux.misc > #1502 > unrolled thread

USB hub issue

Started byClark Smith <noaddress@nowhere.net>
First post2011-06-28 16:17 +0000
Last post2011-07-02 18:18 +0200
Articles 6 — 6 participants

Back to article view | Back to comp.os.linux.misc


Contents

  USB hub issue Clark Smith <noaddress@nowhere.net> - 2011-06-28 16:17 +0000
    Re: USB hub issue Joost Kremers <joostkremers@yahoo.com> - 2011-06-28 17:46 +0000
    Re: USB hub issue King Beowulf <kingbeowulf@nomail.com> - 2011-07-01 19:52 +0000
      Re: USB hub issue Peter Chant <peteRE@MpeteOzilla.Vco.ukE> - 2011-07-02 20:32 +0100
    Re: USB hub issue Todd <Todd@invalid.com> - 2011-07-01 20:11 -0700
      Re: USB hub issue Aragorn <stryder@telenet.be.invalid> - 2011-07-02 18:18 +0200

#1502 — USB hub issue

FromClark Smith <noaddress@nowhere.net>
Date2011-06-28 16:17 +0000
SubjectUSB hub issue
Message-ID<iucur9$6ri$1@news.albasani.net>
	I have a powered 7-port USB hub connected to a Slackware 13.1 
box. Connected to the hub I have five USB external hard drives This setup 
works fine, but there is a small wrinkle: When a particular hard drive 
has not been "used" for a while (more on this below) the next time I try 
and access it it takes several seconds to respond.

	In order to prevent the actual drives from spinning down when 
unused, I have a cron job that touches an empty file in each drive once 
every two minutes - but the issue persists. At this point my guess is 
that it is to do with the actual USB hub and/or the USB driver. Anybody 
have any ideas what may be going on, and how to test things further?

[toc] | [next] | [standalone]


#1504

FromJoost Kremers <joostkremers@yahoo.com>
Date2011-06-28 17:46 +0000
Message-ID<slrnj0k4rm.ec5.joostkremers@j.kremers4.news.arnhem.chello.nl>
In reply to#1502
[Followup-To: header set to alt.os.linux.slackware]
Clark Smith wrote:
> 	In order to prevent the actual drives from spinning down when 
> unused, I have a cron job that touches an empty file in each drive once 
> every two minutes - but the issue persists. At this point my guess is 
> that it is to do with the actual USB hub and/or the USB driver. Anybody 
> have any ideas what may be going on, and how to test things further?

most likely, the cron job just touches the file in the in-memory cache of the
usb drives' file systems.


-- 
Joost Kremers                                      joostkremers@yahoo.com
Selbst in die Unterwelt dringt durch Spalten Licht
EN:SiS(9)

[toc] | [prev] | [next] | [standalone]


#1544

FromKing Beowulf <kingbeowulf@nomail.com>
Date2011-07-01 19:52 +0000
Message-ID<iul8i5$6tf$1@news.albasani.net>
In reply to#1502
On Tue, 28 Jun 2011 16:17:45 +0000, Clark Smith wrote:

> I have a powered 7-port USB hub connected to a Slackware 13.1 box.
> Connected to the hub I have five USB external hard drives This setup
> works fine, but there is a small wrinkle: When a particular hard drive
> has not been "used" for a while (more on this below) the next time I try
> and access it it takes several seconds to respond.
> 
> ---snip---

depending on the USB drive band/model, and depending on the kernel 
version, certain ones either don't or take a long time to wake up from 
power save spin down. This can usually be turned off via sdparm. For 
example, for for 500GB Seagate FreeAgent USB:

# sdparm -al /dev/sdd
# sdparm --clear STANDBY -6 /dev/sdd
# sdparm -al /dev/sdd 

See sdparm manpage, http://sg.danny.cz/sg/sdparm.html and numerous posts 
on linuxquestions.org


-- 
Linux Counter #230211

[toc] | [prev] | [next] | [standalone]


#1550

FromPeter Chant <peteRE@MpeteOzilla.Vco.ukE>
Date2011-07-02 20:32 +0100
Message-ID<nb84e8-gbh.ln1@phoenix.fire>
In reply to#1544
King Beowulf wrote:

> # sdparm -al /dev/sdd
> # sdparm --clear STANDBY -6 /dev/sdd
> # sdparm -al /dev/sdd
> 
> See sdparm manpage, http://sg.danny.cz/sg/sdparm.html and numerous posts
> on linuxquestions.org

Does that work for usb connected drives?  Din't use to but things may have 
changed.

Pete

-- 
http://www.petezilla.co.uk

[toc] | [prev] | [next] | [standalone]


#1546

FromTodd <Todd@invalid.com>
Date2011-07-01 20:11 -0700
Message-ID<ium29d$stt$1@dont-email.me>
In reply to#1502
On 06/28/2011 09:17 AM, Clark Smith wrote:
> 	I have a powered 7-port USB hub connected to a Slackware 13.1
> box. Connected to the hub I have five USB external hard drives This setup
> works fine, but there is a small wrinkle: When a particular hard drive
> has not been "used" for a while (more on this below) the next time I try
> and access it it takes several seconds to respond.
>
> 	In order to prevent the actual drives from spinning down when
> unused, I have a cron job that touches an empty file in each drive once
> every two minutes - but the issue persists. At this point my guess is
> that it is to do with the actual USB hub and/or the USB driver. Anybody
> have any ideas what may be going on, and how to test things further?
>

Guys,

    If I am not mistaken, isn't there something in the /etc/fstab
that turn off caching on your drive so his touch ping will actually
touch the drive?

    Also, how about issuing a "sync" command after the "touch" command?

    I know I am not being much help here.

-T

[toc] | [prev] | [next] | [standalone]


#1548

FromAragorn <stryder@telenet.be.invalid>
Date2011-07-02 18:18 +0200
Message-ID<iungbo$gie$1@dont-email.me>
In reply to#1546
On Saturday 02 July 2011 05:11 in alt.os.linux.slackware, Todd 
enlightened humanity with the following words...:

> On 06/28/2011 09:17 AM, Clark Smith wrote:
>
>> I have a powered 7-port USB hub connected to a Slackware 13.1
>> box. Connected to the hub I have five USB external hard drives This
>> setup works fine, but there is a small wrinkle: When a particular
>> hard drive has not been "used" for a while (more on this below) the
>> next time I try and access it it takes several seconds to respond.
>>
>> In order to prevent the actual drives from spinning down when
>> unused, I have a cron job that touches an empty file in each drive
>> once every two minutes - but the issue persists. At this point my
>> guess is that it is to do with the actual USB hub and/or the USB
>> driver. Anybody have any ideas what may be going on, and how to test
>> things further?
> 
> Guys,
> 
>     If I am not mistaken, isn't there something in the /etc/fstab
> that turn off caching on your drive so his touch ping will actually
> touch the drive?

You are talking of the "sync" mount option - the default for hard disk 
partitions is "async".

>     Also, how about issuing a "sync" command after the "touch"
>     command?

That should also do the trick.  However, the automatic spinning down of 
the hard disks should be disabled using hdparm, sdparm or something of 
the likes.  Probably the latter, since the OP's drives are USB-
connected, and Linux accesses all USB storage peripherals via the SCSI 
midlayer.


-- 
Aragorn
(registered GNU/Linux user #223157)

[toc] | [prev] | [standalone]


Back to top | Article view | comp.os.linux.misc


csiph-web