debian-user-digest Digest V2007 #2294

From: <debian-user-digest-request(at)lists.debian.org>
Date: Sat Sep 01 2007 - 22:18:15 EDT

Content-Type: text/plain

debian-user-digest Digest Volume 2007 : Issue 2294

Today's Topics:

  Help! Update screwed my box up 'auto  [ Martin Waller  ]
  Re: lvm mirror or mdadm mirror or ra  [ Alex Samad  ]
  Re: Amarok + mp3 embedded album art   [ koffiejunkie  ]
  What uses xorg process?               [ Thomas Anderson  ]
  Re: Acceptable CPU temperature range  [ "Douglas A. Tutty"  ]

Date: Sat, 01 Sep 2007 21:32:48 +0100
From: Martin Waller <martinej.waller@ntlworld.com> To: debian-user@lists.debian.org
Subject: Help! Update screwed my box up 'automagically'!

Message-ID: <46D9CC70.8080507@ntlworld.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Hi,

No idea whats gone wrong - I did a get updates thing using dselect whilst installing some other software and find loads of my configurations have changed.

I guess its a result of this: http://www.debian.org/News/2007/20070817

First, my /boot/grub/menu.lst has been dicked around with:

• originally I had two options - multi and single user mode (Debian GNU/Linux, kernel 2.6.18-4-686) now I have four with kernel 2.6.18-5-86 options higher up the list...this stopped my network from working unless I select the now not default option I wanted originally. I remember some operating system called Windows that dicked around with configuration files without asking - I avoid it wherever possible.. Don't tell me Debian has gone this way? *rolls eyes*. Does this mean my kernel has been updated automatically? Please no.

Secondly, OK so I can delete that crap from my menu.lst or select the other oirignal options - but now my X sessions all start up in 640x480 (gnome) and I can't change the resolution!

I'd got all this set up nicely, and a simple update dicks around with stuff and screws up my system - how can this be possible?

I guess I need 'educating' in how Debian operates now, and how to 'undo' the bad side effects I'm suffering, but if I can't resolve this stuff I'll switch distros. I've only just updated from Debian3.1 where I never had such issues - I'm upset and confused.

Thanks for any help,

Martin

Date: Sat, 1 Sep 2007 22:57:21 +0200
From: Matus UHLAR - fantomas <uhlar@fantomas.sk> To: debian-user@lists.debian.org
Subject: Re: [OT] smart reporting trouble

Message-ID: <20070901205721.GA2432@fantomas.sk>
Content-Type: text/plain; charset=iso-8859-2
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On 17.07.07 10:26, Andr=E9 Berger wrote:
> Back from vacation, I found messages from smartd (sarge; 2.6.21.5) in
> root's mailbox:

>=20

>   SMART error (CurrentPendingSector) detected
>     The following warning/error was logged by the smartd daemon:
>     Device: /dev/hda, 1061 Currently unreadable (pending) sectors=20

>=20

>   SMART error (OfflineUncorrectableSector) detected
>     The following warning/error was logged by the smartd daemon:
>     Device: /dev/hda, 1746 Offline uncorrectable sectors

>=20

> The number of unread and uncorrectable sectors seems to be constant
> over two weeks. I'm not sure if I have to replace the HDD?

replace the disk ASAP. I wouldn't trust it.

--=20
Matus UHLAR - fantomas, uhlar(at)fantomas.sk ; http://www.fantomas.sk/ Warning: I wish NOT to receive e-mail advertising to this address. Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu. Eagles may soar, but weasels don't get sucked into jet engines.=20

Date: Sat, 1 Sep 2007 21:17:44 +0000 (UTC) From: Felix Karpfen <felixk@webone.com.au> To: debian-user@lists.debian.org
Subject: Acceptable CPU temperature range of idle computer?

Message-ID: 
Content-Type:  text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Since upgrading from Sarge to Etch, my CPU temperature appears to be running 5-10 deg.C higher; immediately after booting it is a healthy 30 deg.C and it creeps up to 45-50 deg.C within an hour. On the idle (?) computer there are some 120 processes running and the system load is abou= t
40%.

Am I heading for a problem?

The documentation says that the CPU is:

"Socket 478 for Intel4/Celeron"

and I have failed to locate any reference to an acceptable temperature range.

All advice will be gratefully received.

Felix Karpfen

--=20
Felix Karpfen
Public Key 72FDF9DF (DH/DSA)

Date: Sat, 1 Sep 2007 14:24:08 -0700 (PDT) From: Jeff D <fixedored@gmail.com>
To: debian user <debian-user@lists.debian.org> Subject: Re: Help! Update screwed my box up 'automagically'! Message-ID: <Pine.LNX.4.62.0709011402590.3915@proto.technobounce.com> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed

On Sat, 1 Sep 2007, Martin Waller wrote:

> Hi,

>

> No idea whats gone wrong - I did a get updates thing using dselect whilst
> installing some other software and find loads of my configurations have
> changed.

>

> I guess its a result of this: http://www.debian.org/News/2007/20070817

>

> First, my /boot/grub/menu.lst has been dicked around with:
> - originally I had two options - multi and single user mode (Debian
> GNU/Linux, kernel 2.6.18-4-686)
> now I have four with kernel 2.6.18-5-86 options higher up the list...this
> stopped my network from working unless I select the now not default option I
> wanted originally. I remember some operating system called Windows that
> dicked around with configuration files without asking - I avoid it wherever
> possible.. Don't tell me Debian has gone this way? *rolls eyes*.
> Does this mean my kernel has been updated automatically? Please no.

>

> Secondly, OK so I can delete that crap from my menu.lst or select the other
> oirignal options - but now my X sessions all start up in 640x480 (gnome) and
> I can't change the resolution!

>

> I'd got all this set up nicely, and a simple update dicks around with stuff
> and screws up my system - how can this be possible?

>

> I guess I need 'educating' in how Debian operates now, and how to 'undo' the
> bad side effects I'm suffering, but if I can't resolve this stuff I'll switch
> distros. I've only just updated from Debian3.1 where I never had such issues
> - I'm upset and confused.

>

> Thanks for any help,

>
> Martin

First thing, take a deep breath. It will be fine. Second, read this: http://catb.org/~esr/faqs/smart-questions.html. As 'help I just updated my kernel and now stuff is broke', really doesn't help anyone, epseicially you. How are we to advise you on that?

Anyway, did this happen in the transition from sarge to etch? or after the etch upgrade? What kind of network card do you have? Any reason you were doing this through dselect rather than apt-get or aptitude? Maybe even a 'aptitude update && aptitude upgrade' may solve some of these issues for you as well..

For your video problem, you can try :
dpkg-reconfigure xorg-xserver

that might clear that up.

But, in order for anyone to help you, you're going to have to give us some more info.

Thx
Jeff

-+-
8 out of 10 Owners who Expressed a Preference said Their Cats Preferred Techno.

Date: Sun, 2 Sep 2007 08:44:21 +1000
From: Alex Samad <alex@samad.com.au>
To: debian-user@lists.debian.org
Subject: Re: lvm mirror or mdadm mirror or raid 5 Message-ID: <20070901224421.GY17607@samad.com.au> Content-Type: multipart/signed; micalg=pgp-sha1;

        protocol="application/pgp-signature"; boundary="AAkN98o3X3ouhQaz" Content-Disposition: inline

--AAkN98o3X3ouhQaz

Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sat, Sep 01, 2007 at 09:48:38AM -0400, Douglas A. Tutty wrote:
> On Sat, Sep 01, 2007 at 12:16:00PM +1000, Alex Samad wrote:
> > raid 1 /
> > raid 1 /boot
> > raid 5 - pv
> >=20
> > I have been looking at why not to raid 5 and I would go to raid 10 but =
I can=20
> > only put 3 drives in my machine.
> >=20
> > Currently have 3x250G and am looking at getting 3x500G, if I go to mdad=
m raid1,=20
> > then I have 500G safe 500G unsafe which is not really what I want.
> >=20
> > I am going to stick with the /, /boot raid1 sets.
> >=20
> > But I was thinking about setting up the rest of the space as pv's and t=
hem=20
> > using lvm's dm-mirror to manage the mirroring of space ? then if I don=
't want=20
> > any redundancy on my space I can lvcreate without dm-mirror.
> >=20
> > Any one doing this, I haven't had much experience with dm-mirror.
> > my thoughts were=20
> >=20
> > 1 big vgroup
> > and lots of lv's some with mirrors and some without ?

>=20
>=20

> I don't see a mirror option in Etch's LV create.

Okay I have lenny but I thought mirroring had been in for a while, from my = man=20
lvcreate=20

 -m, --mirrors Mirrors

I haven't looked at this previously so I might be wrong.

>=20

> Here's how I would do it:

>=20
> Partition 1: 3-way raid1 64 MB for /boot

I usually go for 500M, leaves me lots of room to leave stuff not sure why 3 way though, the spare space on the 3rd drive if i go 2 way i=  use=20
for tmp or swap or ...

> Partition 2: 3-way raid1 300 MB for /

again not sure why the 3 way, currently i give 10G for this

> Partition 3: 1/2 of remainder on each drive
> Partition 4: the other 1/2 of remainder.

>=20

> Drives are a,b, and c.

>=20

> md0: a1, b1, c1  >> raid1 /boot
> md1: a2, b2, c2  >> raid1 /
> md2: a3, b4      >> raid1, as PV for LVM
> md3: b3, c4      >> raid1, as PV for LVM
> md4: c1, a4	 >> raid1, as PV for LVM

>=20

> This way, you can still loose one drive and be safe. It gives you 750
> GB safe, 0 unsafe.

nice idea to break up into little pv

>=20

> Of course, performance won't be as good as if you had 4 drives so that
> you didn't have more than one busy md per drive, but this is a good
> compromise.

>=20
> Why is it that you can't fit in a 4th drive? Are there only 3 bays?
its a shuttle case, it only has bays for 2 hard drives, I suppose I could= =20
remove the dvd drive, but I use it a lot, so 3 is the magic number, and it= =20
worked well for me, whilst I was a raid 5 guy, but it I am thinking more ba= ck=20
to raid1, but I would like to make the decision to mirror or not at lvcreat= e=20
time not underneath

been playing with some HP eva's and I like being able to set the raid level=  at=20
the lun level

> What about an external bay that operates at SATA speeds; look at
> addonics.

>=20

> Doug.

>=20
>=20

> --=20
> To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org=20
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian=
=2Eorg
>=20
>=20

--AAkN98o3X3ouhQaz

Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFG2etFkZz88chpJ2MRAmzfAKDo57hUMFlWrL9JG8WYTQBFpP1JyQCgogYZ 27EIGwzvQl3/+K0uMD506vs=
=bSoW
-----END PGP SIGNATURE----- --AAkN98o3X3ouhQaz--

Date: Sat, 01 Sep 2007 23:43:16 +0100
From: koffiejunkie <koffiejunkielistlurker@koffiejunkie.za.net> To: debian-user@lists.debian.org
Subject: Re: Amarok + mp3 embedded album art

Message-ID: <46D9EB04.1080309@koffiejunkie.za.net>
Do you need help?
In addition to our free technical library, Pantek provides professional IT services to companies who utilize Open Source software:
Learn more about Pantek's Technical Support offerings
Place an order for 24/7/365 Technical Support here
Learn more about Pantek, who we are, and what we do
Chat with a Live Operator
Related lists.debian.org Links
debian-announce
debian-apache
debian-arm-changes
debian-beowulf
debian-boot
debian-cd
debian-cd-vendors
debian-changes
debian-changes-digest
debian-isp
debian-kde
debian-laptop
debian-mirrors-announce
debian-news
debian-security
debian-security-announce
debian-testing
debian-user
debian-user-digest
Request more information about Pantek
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

koffiejunkie wrote:
> Hi guys,
>
> I'm seeing a strange issue. I was busy sorting my music collection,
> fixing tags with easytag (1.99.13-0.0 from debian-multimedia), and
> adding album art to the files (embedded).
>
> Then, I noticed that Amarok (1.4.4-4) displays garbled image art. I'm
> not enitrely sure if Amarok is feeding me rubbish, or is Easytag not
> saving the files correctly? Amarok displays the embedded album art from
> .m4a files correctly.

Some more info on this - it does appear to be Amarok's fault. I installed Rhythmbox and it was able to read the files and display the album art just fine.

I tried rolling amarok back to the version that comes with etch, but that didn't help. I suspect that whatever amarok uses to read/understand that part of the mp3 file is at fault. Does anyone know what amarok uses?

Thanks

Date: Sun, 02 Sep 2007 01:18:38 +0200
From: Mathias Brodala <info@noctus.net>
To: debian-user@lists.debian.org
Subject: Re: Amarok + mp3 embedded album art Message-ID: <46D9F34E.70801@noctus.net>
Content-Type: multipart/signed; micalg=pgp-sha1;  protocol="application/pgp-signature";
 boundary="------------enig3EB4401A50EE76177C3F9EE9"

This is an OpenPGP/MIME signed message (RFC 2440 and 3156)

--------------enig3EB4401A50EE76177C3F9EE9
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Hi.

koffiejunkie, 02.09.2007 00:43:
> koffiejunkie wrote:

>> Then, I noticed that Amarok (1.4.4-4) displays garbled image art. [=B7= ]

>=20

> [=B7] I suspect that whatever amarok uses to
> read/understand that part of the mp3 file is at fault. Does anyone kno=
w
> what amarok uses?

=46rom "apt-cache show amarok" taglib[0] as it seems.

Regards, Mathias

[0] http://packages.qa.debian.org/t/taglib.html

--=20
debian/rules

--------------enig3EB4401A50EE76177C3F9EE9
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFG2fNWYfUFJ3ewsJgRAj/xAJ96Qeb4L77xd/hdjFbE54zFnfA2GQCff0No LJE65bLbRcxWJUvAV05cmeY=
=KJDL
-----END PGP SIGNATURE----- --------------enig3EB4401A50EE76177C3F9EE9--

Date: Sun, 02 Sep 2007 01:21:12 +0200
From: Thomas Anderson <andersonthomas@gmail.com> To: debian-user@lists.debian.org
Subject: What uses xorg process?

Message-ID: <46D9F3E8.4050500@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to "top", the "xorg" process uses ~25-40% of my AMD 3000+ CPU several times a minute for several seconds at a time. I suspect some other program uses xorg to execute its code for it.

How do I find out what process uses the xorg process to do its stuff for it?

I already tried the graphical top commands tree view to see what subprocesses might use the xorg process. But I could find none.

• -- Regards,

Thomas Anderson
"Quidquid latine dictum sit, altum sonatur"

OpenPGP fingerprint: ED7E 1E98 225A 3FCC 458C B3D7 D625 20E6 F316 BD21 OpenPGP public key: http://todu.dyndns.org/pubkey.txt -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFG2fPo1iUg5vMWvSERAr2YAKCSfXHcyWoJu8R4FZ7ysmXaaMqe4ACeOV7h 2ivwoqZRJzKGFQ++/qVk1SI=
=P3Gd
-----END PGP SIGNATURE-----

Date: Sat, 1 Sep 2007 20:25:09 -0400
From: Carl Fink <carl@finknetwork.com>
To: debian-user@lists.debian.org
Subject: Re: USB ports not responding

Message-ID: <20070902002509.GC22356@nitpicking.com>
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Sat, Sep 01, 2007 at 08:38:00PM +0200, Rickard Lindberg wrote:

> I'm not sure if this problem is hardware or software related. Do you think
> that my USB-ports might have become physically damaged? Is there any way in
> Debian to debug the USB-ports to see what is going on?

It's fairly easy to tell: boot with a Knoppix or other LiveCD and see if USB works.

Debian USB support (at least in Testing) continues to be unreliable. Just Thursday I plugged in my own removable hard drive, which worked fine as /dev/sde. I removed it, used it on another box, then plugged it back in to transfer files ... and no detection, no /dev/sde existed, and no messages in /var/log/messages when I plugged or unplugged it. I tried restarting udev, and even rmmod'ed all USB related modules and re-modprobed them, and it still won't detect the drive.

-- 
Carl Fink                                   nitpicking@nitpicking.com 

Read my blog at nitpickingblog.blogspot.com.  Reviews!  Observations!
Stupid mistakes you can correct!

Date: Sat, 1 Sep 2007 21:13:42 -0400 From: "Douglas A. Tutty" <dtutty@porchlight.ca> To: debian-user@lists.debian.org Subject: Re: Acceptable CPU temperature range of idle computer? Message-ID: <20070902011342.GA13987@titan.hooton> Content-Type: text/plain; charset=us-ascii Content-Disposition: inline On Sat, Sep 01, 2007 at 09:17:44PM +0000, Felix Karpfen wrote:
> Since upgrading from Sarge to Etch, my CPU temperature appears to be
> running 5-10 deg.C higher; immediately after booting it is a healthy 30
> deg.C and it creeps up to 45-50 deg.C within an hour. On the idle (?)
> computer there are some 120 processes running and the system load is about
> 40%.
>
> Am I heading for a problem?
>
> The documentation says that the CPU is:
>
> "Socket 478 for Intel4/Celeron"
>
> and I have failed to locate any reference to an acceptable temperature
> range.
>
> All advice will be gratefully received.
Run top. Do you really have 120 processes actually running? What are they if you consider the computer idle? I can't even load up my P-II lower than about 60% idle momentarily or 80% consistantly. Then again it only has 64 MB ram so to do anything big gets into swapping which gives the CPU pleanty of time to rest :) My Athlon64 never gets above 35 C, my video card's GPU never above 40C. Then again, the internal case temp is ambient == 20 C. What controlls your fans? If you're relying on the OS to do it, then maybe there's a problem. My fans are controlled by the BIOS. Doug.

Date: Sat, 1 Sep 2007 21:31:03 -0400 From: "Douglas A. Tutty" <dtutty@porchlight.ca> To: debian-user@lists.debian.org Subject: Re: lvm mirror or mdadm mirror or raid 5 Message-ID: <20070902013103.GB13987@titan.hooton> Content-Type: text/plain; charset=us-ascii Content-Disposition: inline On Sun, Sep 02, 2007 at 08:44:21AM +1000, Alex Samad wrote:
> On Sat, Sep 01, 2007 at 09:48:38AM -0400, Douglas A. Tutty wrote:
> > On Sat, Sep 01, 2007 at 12:16:00PM +1000, Alex Samad wrote:
> > > raid 1 / raid 1 /boot raid 5 - pv
> > >
> > > I have been looking at why not to raid 5 and I would go to raid 10
> > > but I can only put 3 drives in my machine.
> > >
> > > Currently have 3x250G and am looking at getting 3x500G, if I go to
> > > mdadm raid1, then I have 500G safe 500G unsafe which is not really
> > > what I want.

> > > 1 big vgroup and lots of lv's some with mirrors and some without ?
> >
> > I don't see a mirror option in Etch's LV create.
>
> Okay I have lenny but I thought mirroring had been in for a while,
> from my man lvcreate
>
> -m, --mirrors Mirrors
>
> I haven't looked at this previously so I might be wrong.
Before Etch came out, when I knew that its installer would be doing LVM, I decided that when I moved from Sarge that I'd go with LVM. On Sarge's lvcreate, there was a -m --mirrors option. When Etch came out it wasn't there. I haven't looked into why not but its not there. To me, its the one big thing missing from LMV. I suppose though that if we all start mirroring in LVM instead of setting up md, then we'll need an LV version of mdadm to email us when a mirror degrades from a failing drive.
>
> >
> > Here's how I would do it:
> >
> > Partition 1: 3-way raid1 64 MB for /boot
>
> I usually go for 500M, leaves me lots of room to leave stuff not sure
> why 3 way though, the spare space on the 3rd drive if i go 2 way i use
> for tmp or swap or ...
>
If you don't have swap protected, then a failed drive could crash the system. Put the swap on LVM. You could also put / on LVM but I didn't suggest that since you seemed to want to keep it to a non-LV. I have two drives, 70 GB each. 3 partitions: md0 for /boot, md1 raid1 for LVM VG-mirror for the system stuff and then the two third partitions for VG-cat for non-mirror protected stuff. /home is on a regular LV, /var/tmp is on a striped LV. /tmp is on tmpfs. I'm running amd64 and have a 300 M LV for / with ext3 gives 291 MB total and I have 170 MB used with 107 free with two kernels and their modules installed. I suggested that you go with 3-way raid1 since you have three huge drives anyway, you wouldn't miss 300 MB. Since its a small box with three drives in it, cooling is tight. I would guess that you may have an increased risk of drive failure because of it. With 3-way raid1 for /boot and either three-way raid1 or a mirrord LV if told to mirror to all three drives, your whole system is protected from one drive failure and your / and /boot are protected from two drive failures.
> > Partition 2: 3-way raid1 300 MB for /
> again not sure why the 3 way, currently i give 10G for this
300 MB for / is pleanty if you put all the usual mount points on mirrored LVs.
> nice idea to break up into little pv
> >
> > Of course, performance won't be as good as if you had 4 drives so
> > that you didn't have more than one busy md per drive, but this is a
> > good compromise.
> >
> > Why is it that you can't fit in a 4th drive? Are there only 3 bays?

> its a shuttle case, it only has bays for 2 hard drives, I suppose I
How do you fit in three drives plus a DVD then?
> could remove the dvd drive, but I use it a lot, so 3 is the magic
> number, and it worked well for me, whilst I was a raid 5 guy, but it I
> am thinking more back to raid1, but I would like to make the decision
> to mirror or not at lvcreate time not underneath
>
> been playing with some HP eva's and I like being able to set the raid
> level at the lun level
>
>
> > What about an external bay that operates at SATA speeds; look at
> > addonics.
Doug.

Date: Sat, 1 Sep 2007 21:33:25 -0400 From: "Douglas A. Tutty" <dtutty@porchlight.ca> To: debian-user@lists.debian.org Subject: Re: What uses xorg process? Message-ID: <20070902013325.GC13987@titan.hooton> Content-Type: text/plain; charset=us-ascii Content-Disposition: inline On Sun, Sep 02, 2007 at 01:21:12AM +0200, Thomas Anderson wrote:
> According to "top", the "xorg" process uses ~25-40% of my AMD 3000+ CPU
> several times a minute for several seconds at a time. I suspect some
> other program uses xorg to execute its code for it.
>
> How do I find out what process uses the xorg process to do its stuff for it?
Anything that puts anything on the screen gets xorg to do it. Also, many apps pre-process images (e.g. the next page of a pdf) and store the image in xorg's memory. KDE does this. Its a royal pain when I want to use KPF on my Athlon64 (with pleanty of memory) by sshing in from my P-II with 64 MB. Doug.

Date: Sun, 2 Sep 2007 11:51:04 +1000 From: Alex Samad <alex@samad.com.au> To: debian-user@lists.debian.org Subject: Re: lvm mirror or mdadm mirror or raid 5 Message-ID: <20070902015104.GA28799@samad.com.au> Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="u3/rZRmxL6MmkK24" Content-Disposition: inline --u3/rZRmxL6MmkK24 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Sat, Sep 01, 2007 at 09:31:03PM -0400, Douglas A. Tutty wrote:
> On Sun, Sep 02, 2007 at 08:44:21AM +1000, Alex Samad wrote:
> > On Sat, Sep 01, 2007 at 09:48:38AM -0400, Douglas A. Tutty wrote:
> > > On Sat, Sep 01, 2007 at 12:16:00PM +1000, Alex Samad wrote:
> > > > raid 1 / raid 1 /boot raid 5 - pv
> > > >=20
> > > > I have been looking at why not to raid 5 and I would go to raid 10
> > > > but I can only put 3 drives in my machine.
> > > >=20
> > > > Currently have 3x250G and am looking at getting 3x500G, if I go to
> > > > mdadm raid1, then I have 500G safe 500G unsafe which is not really
> > > > what I want.
> =20
> > > > 1 big vgroup and lots of lv's some with mirrors and some without ?
> > >=20
> > > I don't see a mirror option in Etch's LV create.
> >=20
> > Okay I have lenny but I thought mirroring had been in for a while,
> > from my man lvcreate=20
> >=20
> > -m, --mirrors Mirrors
> >=20
> > I haven't looked at this previously so I might be wrong.
>=20
> Before Etch came out, when I knew that its installer would be doing LVM,
> I decided that when I moved from Sarge that I'd go with LVM. On Sarge's
> lvcreate, there was a -m --mirrors option. When Etch came out it wasn't
> there. I haven't looked into why not but its not there. To me, its the
> one big thing missing from LMV. I suppose though that if we all start
> mirroring in LVM instead of setting up md, then we'll need an LV version
> of mdadm to email us when a mirror degrades from a failing drive.
This is part of no having played with it, was hoping there would have been = a=20 few people who are using it to provide some feedback - yeah its good, nope = its=20 bad. Or even some people to say dm-mirror is not as good/fast as mdadm.. Guess when I get my new drives I am going to have to try it. My problem is= the=20 upgrade path, I don't have the space to backup somewhere else and reload on= =20 fresh disk....... >=20
> >=20
> > >=20
> > > Here's how I would do it:
> > >=20
> > > Partition 1: 3-way raid1 64 MB for /boot
> >=20
> > I usually go for 500M, leaves me lots of room to leave stuff not sure
> > why 3 way though, the spare space on the 3rd drive if i go 2 way i use
> > for tmp or swap or ...
> >=20
>=20
> If you don't have swap protected, then a failed drive could crash the
> system. Put the swap on LVM. You could also put / on LVM but I didn't
> suggest that since you seemed to want to keep it to a non-LV. =20
true >=20
> I have two drives, 70 GB each. 3 partitions: md0 for /boot, md1 raid1
> for LVM VG-mirror for the system stuff and then the two third partitions
> for VG-cat for non-mirror protected stuff. /home is on a regular LV,
> /var/tmp is on a striped LV. /tmp is on tmpfs. I'm running amd64 and
> have a 300 M LV for / with ext3 gives 291 MB total and I have 170 MB
> used with 107 free with two kernels and their modules installed.
>=20
> I suggested that you go with 3-way raid1 since you have three huge
> drives anyway, you wouldn't miss 300 MB. Since its a small box with
> three drives in it, cooling is tight. I would guess that you may have
> an increased risk of drive failure because of it. With 3-way raid1 for
> /boot and either three-way raid1 or a mirrord LV if told to mirror to
> all three drives, your whole system is protected from one drive failure
> and your / and /boot are protected from two drive failures.
>=20
> > > Partition 2: 3-way raid1 300 MB for /
> > again not sure why the 3 way, currently i give 10G for this
>=20
> 300 MB for / is pleanty if you put all the usual mount points on
> mirrored LVs.
>=20
> > nice idea to break up into little pv
> > >=20
> > > Of course, performance won't be as good as if you had 4 drives so
> > > that you didn't have more than one busy md per drive, but this is a
> > > good compromise.
> > >=20
> > > Why is it that you can't fit in a 4th drive? Are there only 3 bays?
>=20
> > its a shuttle case, it only has bays for 2 hard drives, I suppose I
>=20
> How do you fit in three drives plus a DVD then?
the shuttle version i have has room for 3 internal + 1 dvd, I was saying 1= =20 could remove the dvd and install another drive, but i use the dvd too much >=20
> > could remove the dvd drive, but I use it a lot, so 3 is the magic
> > number, and it worked well for me, whilst I was a raid 5 guy, but it I
> > am thinking more back to raid1, but I would like to make the decision
> > to mirror or not at lvcreate time not underneath
> >=20
> > been playing with some HP eva's and I like being able to set the raid
> > level at the lun level
> >=20
> >=20
> > > What about an external bay that operates at SATA speeds; look at
> > > addonics.
i work has a much bigger budget than mine, i was look at 500G samsungs driv= es=20 for $110 (AUS$), which was a good deal >=20
> Doug.
>=20 >=20
> --=20
> To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org=20
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian=
=2Eorg >=20 >=20 --u3/rZRmxL6MmkK24 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: Digital signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFG2hcIkZz88chpJ2MRAnDAAKC6KVF6iYLi/wvZQXOq27p1h6qHWgCdF4zJ Qvn3uw6Ato0pZLDLaVk3hCs= =RN/N -----END PGP SIGNATURE----- --u3/rZRmxL6MmkK24-- End of debian-user-digest Digest V2007 Issue #2294 ************************************************** Received on Sat Sep 1 22:11:59 2007