[gentoo-user] A little light relief from endless problems

[gentoo-user] A little light relief from endless problems

[Peter Humphrey]

As it's Sunday, here's an odd little thing.

Not long ago, while booting this machine, four ext3 partitions needed checks 
on remount count reaching zero. They had been set to 23, 24, 25 and 26 
mounts. (I didn't choose the numbers; they were allocated at the time I was 
creating the file system.)

Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 = 358,800 
times all told. At, say, two reboots per day, that would take rather a long 
time: a little under 500 years if my arithmetic is working.

-- 
Rgds
Peter

[gentoo-user] Re: A little light relief from endless problems

[Nikos Chantziaras]

Peter Humphrey wrote:
> As it's Sunday, here's an odd little thing.
> 
> Not long ago, while booting this machine, four ext3 partitions needed checks 
> on remount count reaching zero. They had been set to 23, 24, 25 and 26 
> mounts. (I didn't choose the numbers; they were allocated at the time I was 
> creating the file system.)
> 
> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 = 358,800 
> times all told. At, say, two reboots per day, that would take rather a long 
> time: a little under 500 years if my arithmetic is working.

I think you're confused.  23 means a check each 23 mounts.  With 2 
mounts per day, that's a check every 12 days for the first and second disk.

Also, except mount count, there's also a time-based check.  The check 
happens whichever of the two expires first (otherwise, a system that 
gets rebooted once each two months or such would get checked in a timely 
manner.)

Re: [gentoo-user] Re: A little light relief from endless problems

[Allan Gottlieb]

At Sun, 01 Mar 2009 13:57:30 +0200 Nikos Chantziaras <realnc@arcor.de> wrote:

> Peter Humphrey wrote:
>> As it's Sunday, here's an odd little thing.
>>
>> Not long ago, while booting this machine, four ext3 partitions
>> needed checks on remount count reaching zero. They had been set to
>> 23, 24, 25 and 26 mounts. (I didn't choose the numbers; they were
>> allocated at the time I was creating the file system.)
>>
>> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 =
>> 358,800 times all told. At, say, two reboots per day, that would
>> take rather a long time: a little under 500 years if my arithmetic
>> is working.
>
> I think you're confused.  23 means a check each 23 mounts.  With 2
> mounts per day, that's a check every 12 days for the first and second
> disk.

I think the point is that 23, 24, 25, 26 are relatively prime so that,
if N is initially zero, it takes 23x24x25x26 increments initially for (N
mod 23), ..., (N mod 26) to all again be zero.

> Also, except mount count, there's also a time-based check.  The check
> happens whichever of the two expires first (otherwise, a system that
> gets rebooted once each two months or such would get checked in a
> timely manner.)

This second point is quite valid.

allan

[gentoo-user] Re: A little light relief from endless problems

[Nikos Chantziaras]

Allan Gottlieb wrote:
> At Sun, 01 Mar 2009 13:57:30 +0200 Nikos Chantziaras <realnc@arcor.de> wrote:
> 
>> Peter Humphrey wrote:
>>> As it's Sunday, here's an odd little thing.
>>>
>>> Not long ago, while booting this machine, four ext3 partitions
>>> needed checks on remount count reaching zero. They had been set to
>>> 23, 24, 25 and 26 mounts. (I didn't choose the numbers; they were
>>> allocated at the time I was creating the file system.)
>>>
>>> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 =
>>> 358,800 times all told. At, say, two reboots per day, that would
>>> take rather a long time: a little under 500 years if my arithmetic
>>> is working.
>> I think you're confused.  23 means a check each 23 mounts.  With 2
>> mounts per day, that's a check every 12 days for the first and second
>> disk.
> 
> I think the point is that 23, 24, 25, 26 are relatively prime so that,
> if N is initially zero, it takes 23x24x25x26 increments initially for (N
> mod 23), ..., (N mod 26) to all again be zero.

Why would it matter if they're all zero or not?

Re: [gentoo-user] Re: A little light relief from endless problems

[Allan Gottlieb]

At Sun, 01 Mar 2009 16:27:17 +0200 Nikos Chantziaras <realnc@arcor.de> wrote:

> Allan Gottlieb wrote:
>> At Sun, 01 Mar 2009 13:57:30 +0200 Nikos Chantziaras <realnc@arcor.de> wrote:
>>
>>> Peter Humphrey wrote:
>>>> As it's Sunday, here's an odd little thing.
>>>>
>>>> Not long ago, while booting this machine, four ext3 partitions
>>>> needed checks on remount count reaching zero. They had been set to
>>>> 23, 24, 25 and 26 mounts. (I didn't choose the numbers; they were
>>>> allocated at the time I was creating the file system.)
>>>>
>>>> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 =
>>>> 358,800 times all told. At, say, two reboots per day, that would
>>>> take rather a long time: a little under 500 years if my arithmetic
>>>> is working.
>>> I think you're confused.  23 means a check each 23 mounts.  With 2
>>> mounts per day, that's a check every 12 days for the first and second
>>> disk.
>>
>> I think the point is that 23, 24, 25, 26 are relatively prime so that,
>> if N is initially zero, it takes 23x24x25x26 increments initially for (N
>> mod 23), ..., (N mod 26) to all again be zero.
>
> Why would it matter if they're all zero or not?

When they are all zero they fsck is triggered.  So it was not surprising
that all the fsck's occurred.  What was surprising is that they all
triggered at the same reboot.

allan

[gentoo-user] Re: A little light relief from endless problems

[Nikos Chantziaras]

Allan Gottlieb wrote:
> At Sun, 01 Mar 2009 16:27:17 +0200 Nikos Chantziaras <realnc@arcor.de> wrote:
> 
>> Allan Gottlieb wrote:
>>> I think the point is that 23, 24, 25, 26 are relatively prime so that,
>>> if N is initially zero, it takes 23x24x25x26 increments initially for (N
>>> mod 23), ..., (N mod 26) to all again be zero.
>> Why would it matter if they're all zero or not?
> 
> When they are all zero they fsck is triggered.  So it was not surprising
> that all the fsck's occurred.  What was surprising is that they all
> triggered at the same reboot.

You might have umounted/mounted them without rebooting.  Or the maximum 
time allowed without a fsck ran out :P

Re: [gentoo-user] A little light relief from endless problems

[KH]

Peter Humphrey schrieb:
> As it's Sunday, here's an odd little thing.
>
> Not long ago, while booting this machine, four ext3 partitions needed checks 
> on remount count reaching zero. They had been set to 23, 24, 25 and 26 
> mounts. (I didn't choose the numbers; they were allocated at the time I was 
> creating the file system.)
>
> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 = 358,800 
> times all told. At, say, two reboots per day, that would take rather a long 
> time: a little under 500 years if my arithmetic is working.
>
>   
Hi,

this is incorrect. 179400 mounts would be enough (24 and 26 can both be
divided by 2). Have you been running gentoo for 250 years? ;-)
Maybe your counts are incorrect in another way. Let's say you install
your system. You create root partition and install, reboot. Than you
install all the other partitions. Maybe something went wrong and you
needed to remount partition 4 four times ...

Hope this helped a little.

kh

Re: [gentoo-user] A little light relief from endless problems

[Allan Gottlieb]

At Sun, 01 Mar 2009 17:49:36 +0100 KH <gentoo-user@konstantinhansen.de> wrote:

> Peter Humphrey schrieb:
>> As it's Sunday, here's an odd little thing.
>>
>> Not long ago, while booting this machine, four ext3 partitions needed checks 
>> on remount count reaching zero. They had been set to 23, 24, 25 and 26 
>> mounts. (I didn't choose the numbers; they were allocated at the time I was 
>> creating the file system.)
>>
>> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 = 358,800 
>> times all told. At, say, two reboots per day, that would take rather a long 
>> time: a little under 500 years if my arithmetic is working.
>>
>>   
> Hi,
>
> this is incorrect. 179400 mounts would be enough (24 and 26 can both be
> divided by 2).

Correct.  I erred in saying that 23,24,25,26 are relatively prime as you
noted.  In general if it was a1,a2,...an the answer would be
LCM(a1,a2,...,an), where LCM abbreviates Least Common Multiple.

allan

Re: [gentoo-user] A little light relief from endless problems

[Momesso Andrea]

[-- Attachment #1: Type: text/plain, Size: 1310 bytes --]

On Sun, Mar 01, 2009 at 02:03:28PM -0500, Allan Gottlieb wrote:
> At Sun, 01 Mar 2009 17:49:36 +0100 KH <gentoo-user@konstantinhansen.de> wrote:
> 
> > Peter Humphrey schrieb:
> >> As it's Sunday, here's an odd little thing.
> >>
> >> Not long ago, while booting this machine, four ext3 partitions needed checks 
> >> on remount count reaching zero. They had been set to 23, 24, 25 and 26 
> >> mounts. (I didn't choose the numbers; they were allocated at the time I was 
> >> creating the file system.)
> >>
> >> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 = 358,800 
> >> times all told. At, say, two reboots per day, that would take rather a long 
> >> time: a little under 500 years if my arithmetic is working.
> >>
> >>   
> > Hi,
> >
> > this is incorrect. 179400 mounts would be enough (24 and 26 can both be
> > divided by 2).
> 
> Correct.  I erred in saying that 23,24,25,26 are relatively prime as you
> noted.  In general if it was a1,a2,...an the answer would be
> LCM(a1,a2,...,an), where LCM abbreviates Least Common Multiple.
> 
> allan

What about battery? If that's a laptop checks are deferred if running on
battery at boot time, so it can happen that all the partitions are
fscked the first time you boot on AC.

=======
TopperH
=======

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: [gentoo-user] A little light relief from endless problems

[Peter Humphrey]

On Monday 02 March 2009 07:09:31 Momesso Andrea wrote:
> On Sun, Mar 01, 2009 at 02:03:28PM -0500, Allan Gottlieb wrote:
> > At Sun, 01 Mar 2009 17:49:36 +0100 KH <gentoo-user@konstantinhansen.de> 
wrote:
> > > Peter Humphrey schrieb:
> > >> As it's Sunday, here's an odd little thing.
> > >>
> > >> Not long ago, while booting this machine, four ext3 partitions
> > >> needed checks on remount count reaching zero. They had been set to
> > >> 23, 24, 25 and 26 mounts. (I didn't choose the numbers; they were
> > >> allocated at the time I was creating the file system.)
> > >>
> > >> Now, this box does get rebooted, but hardly 23 x 24 x 25 x 26 =
> > >> 358,800 times all told. At, say, two reboots per day, that would
> > >> take rather a long time: a little under 500 years if my arithmetic
> > >> is working.
> > >
> > > Hi,
> > >
> > > this is incorrect. 179400 mounts would be enough (24 and 26 can both
> > > be divided by 2).

I wasn't sure how to allow for this common factor, so I ignored it. Thanks 
for the clarification. And no, 250 years is not a lot more credible than 
500! So you might say my error was not significant.

> > Correct.  I erred in saying that 23,24,25,26 are relatively prime as
> > you noted.  In general if it was a1,a2,...an the answer would be
> > LCM(a1,a2,...,an), where LCM abbreviates Least Common Multiple.
> >
> > allan

Nicely  put.

> What about battery? If that's a laptop checks are deferred if running on
> battery at boot time, so it can happen that all the partitions are
> fscked the first time you boot on AC.

It isn't a laptop. It's an ordinary desktop box (or a workstation, depending 
on your point of view: it was sold as a workstation - and it certainly 
makes enough noise for one).

And the other contribution, about manual mounting, is also a red herring. If 
anyone thinks I contributed a hundred thousand operations to the total, 
they need to spend a little time in a quiet room  :-)

-- 
Rgds
Peter

Re: [gentoo-user] A little light relief from endless problems

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 403 bytes --]

On Mon, 2 Mar 2009 10:52:54 +0000, Peter Humphrey wrote:

> And the other contribution, about manual mounting, is also a red
> herring. If anyone thinks I contributed a hundred thousand operations
> to the total, they need to spend a little time in a quiet room  :-)

Six manual mounts would be sufficient.


-- 
Neil Bothwick

Support bacteria - they're the only culture some people have.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]