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


Groups > linux.debian.user > #159587 > unrolled thread

Debian 8.2 RAID conf. problem

Started byerdal daghan <erdaldaghan@gmail.com>
First post2015-10-21 18:30 +0200
Last post2015-10-23 19:00 +0200
Articles 14 — 11 participants

Back to article view | Back to linux.debian.user


Contents

  Debian 8.2 RAID conf. problem erdal daghan <erdaldaghan@gmail.com> - 2015-10-21 18:30 +0200
    Re: Debian 8.2 RAID conf. problem Mario Castelán Castro  <marioxcc.MT@yandex.com> - 2015-10-21 18:40 +0200
      Re: Debian 8.2 RAID conf. problem Tony van der Hoff <tony@vanderhoff.org> - 2015-10-21 18:50 +0200
        Re: Debian 8.2 RAID conf. problem Darac Marjal <mailinglist@darac.org.uk> - 2015-10-22 10:50 +0200
        Re: Debian 8.2 RAID conf. problem Dan Ritter <dsr@randomstring.org> - 2015-10-22 19:00 +0200
      Re: Debian 8.2 RAID conf. problem Pascal Hambourg <pascal@plouf.fr.eu.org> - 2015-10-21 21:20 +0200
      Re: Debian 8.2 RAID conf. problem Cindy-Sue Causey <butterflybytes@gmail.com> - 2015-10-22 16:30 +0200
        Re: Debian 8.2 RAID conf. problem Pascal Hambourg <pascal@plouf.fr.eu.org> - 2015-10-22 21:30 +0200
          Re: Debian 8.2 RAID conf. problem Curt <curty@free.fr> - 2015-10-23 11:00 +0200
            Re: Debian 8.2 RAID conf. problem Chris Bannister <cbannister@slingshot.co.nz> - 2015-10-23 13:50 +0200
              Re: Debian 8.2 RAID conf. problem Lisi Reisz <lisi.reisz@gmail.com> - 2015-10-23 18:50 +0200
                Re: Debian 8.2 RAID conf. problem John Hasler <jhasler@newsguy.com> - 2015-10-23 19:00 +0200
                  Re: Debian 8.2 RAID conf. problem Curt <curty@free.fr> - 2015-10-23 19:30 +0200
                Re: Debian 8.2 RAID conf. problem Curt <curty@free.fr> - 2015-10-23 19:00 +0200

#159587 — Debian 8.2 RAID conf. problem

Fromerdal daghan <erdaldaghan@gmail.com>
Date2015-10-21 18:30 +0200
SubjectDebian 8.2 RAID conf. problem
Message-ID<qm4r4-8qy-85@gated-at.bofh.it>

[Multipart message — attachments visible in raw view] — view raw

Hi,

I have a problem with install debian. When I trying to install the Debian
8.2 with RAID1 configuration it does NOT add bootable flag in guided
partitioning mode. It Doesn't on option.

Please help me.. I have two ssd disks(240gb). I want to make RAID1
configuration.



Erdal DAĞHAN

[toc] | [next] | [standalone]


#159589

FromMario Castelán Castro <marioxcc.MT@yandex.com>
Date2015-10-21 18:40 +0200
Message-ID<qm4AG-at-13@gated-at.bofh.it>
In reply to#159587
El 21/10/15 a las 11:11, erdal daghan escribió:
> Hi,
>
> I have a problem with install debian. When I trying to install the Debian
> 8.2 with RAID1 configuration it does NOT add bootable flag in guided
> partitioning mode. It Doesn't on option.

The problem seems to be your perception, not Debian. Why do you think 
you need that flag?. As far as I know, GRUB ignores it, so it makes no 
difference regarding booting the system. Also, if you use guided 
partitioning, you can manually adjust the suggested partitioning scheme 
before proceeding, so if you really need it, you can easily toggle it.

El 21/10/15 a las 11:11, erdal daghan escribió:
> Please help me.. I have two ssd disks(240gb). I want to make RAID1
> configuration.

You probably refer to ordinary solid state drives which are NOT disks.

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


#159593

FromTony van der Hoff <tony@vanderhoff.org>
Date2015-10-21 18:50 +0200
Message-ID<qm4Kn-nt-35@gated-at.bofh.it>
In reply to#159589
On 21/10/15 17:35, Mario Castelán Castro wrote:
> El 21/10/15 a las 11:11, erdal daghan escribió:
>> Hi,
>>
>> I have a problem with install debian. When I trying to install the Debian
>> 8.2 with RAID1 configuration it does NOT add bootable flag in guided
>> partitioning mode. It Doesn't on option.
>
> The problem seems to be your perception, not Debian. Why do you think
> you need that flag?. As far as I know, GRUB ignores it, so it makes no
> difference regarding booting the system. Also, if you use guided
> partitioning, you can manually adjust the suggested partitioning scheme
> before proceeding, so if you really need it, you can easily toggle it.
>
> El 21/10/15 a las 11:11, erdal daghan escribió:
>> Please help me.. I have two ssd disks(240gb). I want to make RAID1
>> configuration.
>
> You probably refer to ordinary solid state drives which are NOT disks.
>

This raises a question in my mind: Should SSDs be used in RAID arrays?


-- 
Tony van der Hoff        | mailto:tony@vanderhoff.org
Buckinghamshire, England |

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


#159630

FromDarac Marjal <mailinglist@darac.org.uk>
Date2015-10-22 10:50 +0200
Message-ID<qmjJn-5wE-17@gated-at.bofh.it>
In reply to#159593

[Multipart message — attachments visible in raw view] — view raw

On Wed, Oct 21, 2015 at 05:46:55PM +0100, Tony van der Hoff wrote:

>On 21/10/15 17:35, Mario Castelán Castro wrote:
>
>>El 21/10/15 a las 11:11, erdal daghan escribió:
>>
[cut]
>>
>>El 21/10/15 a las 11:11, erdal daghan escribió:
>>
>>>Please help me.. I have two ssd disks(240gb). I want to make RAID1 
>>>configuration.
>>
>>You probably refer to ordinary solid state drives which are NOT disks.
>
>
>This raises a question in my mind: Should SSDs be used in RAID arrays?
>

I don't see a problem with it. RAID is a method to overcome the 
deficiencies of inexpensive disks. With traditional disks, one of those 
major deficiencies is access time, so spreading your data across 
multiple disks may improve access time by allowing you to read from a 
different disk, rather than waiting for a single slow disk.

SSD has good access time, but it deficiencies are size and possibly bulk 
throughput. In terms of size, RAID0 will allow you to combine several 
small SSDs into a single large logical drive. RAID0 would also allow for 
faster throughput than might be achievable with a single disk.

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


#159646

FromDan Ritter <dsr@randomstring.org>
Date2015-10-22 19:00 +0200
Message-ID<qmrnB-8pH-17@gated-at.bofh.it>
In reply to#159593
On Wed, Oct 21, 2015 at 05:46:55PM +0100, Tony van der Hoff wrote:
> On 21/10/15 17:35, Mario Castelán Castro wrote:
> >El 21/10/15 a las 11:11, erdal daghan escribió:
> >>Hi,
> >>
> >>I have a problem with install debian. When I trying to install the Debian
> >>8.2 with RAID1 configuration it does NOT add bootable flag in guided
> >>partitioning mode. It Doesn't on option.
> >
> >The problem seems to be your perception, not Debian. Why do you think
> >you need that flag?. As far as I know, GRUB ignores it, so it makes no
> >difference regarding booting the system. Also, if you use guided
> >partitioning, you can manually adjust the suggested partitioning scheme
> >before proceeding, so if you really need it, you can easily toggle it.
> >
> >El 21/10/15 a las 11:11, erdal daghan escribió:
> >>Please help me.. I have two ssd disks(240gb). I want to make RAID1
> >>configuration.
> >
> >You probably refer to ordinary solid state drives which are NOT disks.
> >
> 
> This raises a question in my mind: Should SSDs be used in RAID arrays?

As always, it depends on what you want to do.

RAID1 is done as a mechanism to give you a window between disk
failure and the need to replace the disk. Using identical SSDs
might not be a good strategy here, because identical write
patterns might produce identical failures.

I would consider something that keeps the secondary disk less
used than the primary - for instance, rsyncing primary to
secondary every so often. Once a day at minimum, once every 30 
minutes or so max.

On the other hand, a database server might make good use of a
RAID10 of two kinds of similar but not identical disks - a disk 
controller on a PCIe x8 slot can be saturated by 16 SATA3 SSDs
all running full-out.

Or you might want to use SSDs as caches for large spinning
disks. Generically, you can do this with bcache; if you use ZFS,
ZIL and L2ARC are what you are looking for.

-dsr-

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


#159606

FromPascal Hambourg <pascal@plouf.fr.eu.org>
Date2015-10-21 21:20 +0200
Message-ID<qm75w-3ST-31@gated-at.bofh.it>
In reply to#159589
Mario Castelán Castro a écrit :
> El 21/10/15 a las 11:11, erdal daghan escribió:
>>
>> I have a problem with install debian. When I trying to install the Debian
>> 8.2 with RAID1 configuration it does NOT add bootable flag in guided
>> partitioning mode. It Doesn't on option.
> 
> The problem seems to be your perception, not Debian. Why do you think 
> you need that flag?. As far as I know, GRUB ignores it

Indeed, but GRUB is not the only piece of software involved in the boot
process. I know by experience that some broken BIOS or UEFI firmwares
require that a primary partition have the boot flag set. Also, the
standard boot code in the MBR requires it too, so if you install GRUB's
boot image in a partition instead of in the MBR, you need to set the
boot flag on that partition.

> Also, if you use guided 
> partitioning, you can manually adjust the suggested partitioning scheme 
> before proceeding, so if you really need it, you can easily toggle it.

Or you can open one of the shells of the Debian installer and run
parted, fdisk or cfdisk.

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


#159642

FromCindy-Sue Causey <butterflybytes@gmail.com>
Date2015-10-22 16:30 +0200
Message-ID<qmp2r-56v-35@gated-at.bofh.it>
In reply to#159589
On 10/21/15, Mario Castelán Castro <marioxcc.MT@yandex.com> wrote:
> El 21/10/15 a las 11:11, erdal daghan escribió:
>> Hi,
>>
>> I have a problem with install debian. When I trying to install the Debian
>> 8.2 with RAID1 configuration it does NOT add bootable flag in guided
>> partitioning mode. It Doesn't on option.
>
> The problem seems to be your perception, not Debian. Why do you think
> you need that flag?. As far as I know, GRUB ignores it, so it makes no
> difference regarding booting the system. Also, if you use guided
> partitioning, you can manually adjust the suggested partitioning scheme
> before proceeding, so if you really need it, you can easily toggle it.


Because it's there as an option in Gparted, I consciously, *manually*
apply the boot flag to a favored partition as soon as I've completed
the initial partition reconfiguring on a hard drive. SO FAR.... the
only place I've ever seen the boot flag matter was with respect to
rerunning GRUB after a newly (debootstrap) installed distro. It's
actually one of those topics I'd experienced and thought about
starting a thread on soon after in case others had feedback for
newcomers.

Was only a couple months ago that I found out that update-grub was at
least one thing about the boot flag that *appeared* to matter, at
least *IN MY CASE* *that day*. After a very long afternoon of not
understanding why GRUB kept reverting to its old config setup, I
accidentally discovered that changing that boot flag from one
partition to another one created instant success on the next
update-grub run I did.

Prior to stumbling on that factor, I'd gone so far as to take the
chance of purging and then reinstalling GRUB hoping it would somehow
cleanse the system (not to mention hoping the hard drive didn't
suddenly fritz in those few seconds it was GRUB-less). What I had set
up then actually let me reinstall GRUB into.. is it MBR (?) which I
had *thought MIGHT* be the problem at that time. And still update-grub
failed. Wasn't until I changed that boot flag that there was
success.... immediate success for whatever reason known to the
software....... :)

Outside of that, nothing I've ever done to date has openly been
reflected as being somehow dependent on that flag. For the record, I
*do* make that change on purpose just because I did waste a LOT of
time that afternoon. Many months ago there was a thread about
longevity of hard drives related to how many boots in their lifetimes.
THAT THREAD came to mind THAT AFTERNOON.... I ate a few boots off my
hard drive's lifetime supply that day. *grin*

Cindy :)

-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* runs with duct tape *

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


#159651

FromPascal Hambourg <pascal@plouf.fr.eu.org>
Date2015-10-22 21:30 +0200
Message-ID<qmtIK-3vR-17@gated-at.bofh.it>
In reply to#159642
Cindy-Sue Causey a écrit :
> 
> Was only a couple months ago that I found out that update-grub was at
> least one thing about the boot flag that *appeared* to matter, at
> least *IN MY CASE* *that day*. After a very long afternoon of not
> understanding why GRUB kept reverting to its old config setup, I
> accidentally discovered that changing that boot flag from one
> partition to another one created instant success on the next
> update-grub run I did.

You obviously misinterpreted what happened. GRUB as a whole ignores the
boot flag. My guess is there was a different GRUB on each partition, and
update-grub applied to the inactive one, so changing the boot flag just
changed which GRUB was booted.

> Many months ago there was a thread about
> longevity of hard drives related to how many boots in their lifetimes.

Nonsense. If the boot count mattered for hard disk longevity, there
would be a SMART attribute for it. There is none.

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


#159662

FromCurt <curty@free.fr>
Date2015-10-23 11:00 +0200
Message-ID<qmGmD-4MY-17@gated-at.bofh.it>
In reply to#159651
On 2015-10-22, Pascal Hambourg <pascal@plouf.fr.eu.org> wrote:
>
>> Many months ago there was a thread about
>> longevity of hard drives related to how many boots in their lifetimes.
>
> Nonsense. If the boot count mattered for hard disk longevity, there
> would be a SMART attribute for it. There is none.
>

There exists a similar "fallacy" concerning light bulbs (if we're
talking hard (re)boot, change of phase).

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


#159668

FromChris Bannister <cbannister@slingshot.co.nz>
Date2015-10-23 13:50 +0200
Message-ID<qmJ19-lc-29@gated-at.bofh.it>
In reply to#159662
On Fri, Oct 23, 2015 at 08:52:42AM +0000, Curt wrote:
> On 2015-10-22, Pascal Hambourg <pascal@plouf.fr.eu.org> wrote:
> >
> >> Many months ago there was a thread about
> >> longevity of hard drives related to how many boots in their lifetimes.
> >
> > Nonsense. If the boot count mattered for hard disk longevity, there
> > would be a SMART attribute for it. There is none.
> >
> 
> There exists a similar "fallacy" concerning light bulbs (if we're
> talking hard (re)boot, change of phase).

fallacy?

I do know that turning incandescent light bulbs off and on shortens their life.
Whether the savings in electricity is worth the cost of the bulb is, of
course, another matter.

-- 
"If you're not careful, the newspapers will have you hating the people
who are being oppressed, and loving the people who are doing the 
oppressing." --- Malcolm X

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


#159686

FromLisi Reisz <lisi.reisz@gmail.com>
Date2015-10-23 18:50 +0200
Message-ID<qmNHs-762-9@gated-at.bofh.it>
In reply to#159668
On Friday 23 October 2015 12:42:07 Chris Bannister wrote:
> On Fri, Oct 23, 2015 at 08:52:42AM +0000, Curt wrote:
> > On 2015-10-22, Pascal Hambourg <pascal@plouf.fr.eu.org> wrote:
> > >> Many months ago there was a thread about
> > >> longevity of hard drives related to how many boots in their lifetimes.
> > >
> > > Nonsense. If the boot count mattered for hard disk longevity, there
> > > would be a SMART attribute for it. There is none.
> >
> > There exists a similar "fallacy" concerning light bulbs (if we're
> > talking hard (re)boot, change of phase).
>
> fallacy?
>
> I do know that turning incandescent light bulbs off and on shortens their
> life. Whether the savings in electricity is worth the cost of the bulb is,
> of course, another matter.

And CFL surely?  If turned on and off too frequently too close together - 
anyhow the older ones?  I don't like them, so haven't used them much.

Lisi

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


#159687

FromJohn Hasler <jhasler@newsguy.com>
Date2015-10-23 19:00 +0200
Message-ID<qmNR9-7hz-29@gated-at.bofh.it>
In reply to#159686
Lisi writes:
> And CFL surely?  If turned on and off too frequently too close
> together - anyhow the older ones?  I don't like them, so haven't used
> them much.

The life of a CFL is pretty much unrelated to how often it is turned on
and off.
-- 
John Hasler 
jhasler@newsguy.com
Elmwood, WI USA

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


#159691

FromCurt <curty@free.fr>
Date2015-10-23 19:30 +0200
Message-ID<qmOka-84O-15@gated-at.bofh.it>
In reply to#159687
On 2015-10-23, John Hasler <jhasler@newsguy.com> wrote:
> Lisi writes:
>> And CFL surely?  If turned on and off too frequently too close
>> together - anyhow the older ones?  I don't like them, so haven't used
>> them much.
>
> The life of a CFL is pretty much unrelated to how often it is turned on
> and off.

http://energy.gov/energysaver/when-turn-your-lights

 The operating life of CFLs is more affected by the number of times they
 are switched on and off. You can generally extend the life of a CFL
 bulb more by switching it on and off less frequently than if you simply
 use it less.

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


#159688

FromCurt <curty@free.fr>
Date2015-10-23 19:00 +0200
Message-ID<qmNR9-7hz-47@gated-at.bofh.it>
In reply to#159686
On 2015-10-23, Lisi Reisz <lisi.reisz@gmail.com> wrote:
>> > >
>> > > Nonsense. If the boot count mattered for hard disk longevity, there
>> > > would be a SMART attribute for it. There is none.
>> >
>> > There exists a similar "fallacy" concerning light bulbs (if we're
>> > talking hard (re)boot, change of phase).
>>
>> fallacy?
>>
>> I do know that turning incandescent light bulbs off and on shortens their
>> life. Whether the savings in electricity is worth the cost of the bulb is,
>> of course, another matter.
>
> And CFL surely?  If turned on and off too frequently too close together - 
> anyhow the older ones?  I don't like them, so haven't used them much.
>
> Lisi
>
>

https://en.wikipedia.org/wiki/Scare_quotes

I am in the dark about these matters.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.debian.user


csiph-web