Re: Cannot get systemd to forget about swap space on a failed disk

[David Wright]

On Tue 30 Jun 2020 at 22:04:02 (-0700), Bob McGowan wrote:
> On 6/29/2020 11:37 PM, Reco wrote:
> > On Mon, Jun 29, 2020 at 11:10:44PM -0700, Bob McGowan wrote:
> > > But I cannot figure out where this might be, or even if this is the 
> > > correct interpretation.
> > Check out the contents of /etc/systemd/system first.
> > Rebuild initramfs second.
> 
> It seems my troubleshooting skills are not up to snuff. :(
> 
> I isolated the wrong disk.  This is partly due to the complexity
> (excessive, I think) of the systemd system, partly due to the fact
> that the startup reports of running jobs associated with disks
> overwrite each other, and I just grabbed the wrong one.
> 
> Regardless, upon commenting the correct disk out of fstab, and adding
> back my new secondary swap, everything works again as expected.
> 
> It may help others to know what I did.
> 
> Part of my original failure to find anything was because I was focused
> on the /lib/systemd, /usr/lib/systemd and /etc/systemd directories. 
> Even though I had read the documentation that referenced /run/systemd,
> I had not checked it out.
> 
> When I did, I immediately found the correct UUID information for the
> actual failing disk.  And the rest is history.

I've never consulted those directories, but just use the files
/etc/fstab and /etc/crypttab (written by yourself), a listing
of /dev/disk, and the appropriate commands for handling swap.
Also lsblk gives a neat overview.

So I would stop your broken partition with:

# swapoff foo
# swapon --show (to see the remaining one and check it out)
# nano /etc/fstab   (and comment the bad entry)
# systemctl daemon-reload

rather than editing fstab while the partition is in use.

Cheers,
David.

Re: Cannot get systemd to forget about swap space on a failed disk

[Bob McGowan]

On 6/29/2020 11:37 PM, Reco wrote:

Hi.

On Mon, Jun 29, 2020 at 11:10:44PM -0700, Bob McGowan wrote:

But I cannot figure out where this might be, or even if this is the correct 
interpretation.

Check out the contents of /etc/systemd/system first.
Rebuild initramfs second.

Reco


Hi,

It seems my troubleshooting skills are not up to snuff. :(

I isolated the wrong disk.  This is partly due to the complexity 
(excessive, I think) of the systemd system, partly due to the fact that 
the startup reports of running jobs associated with disks overwrite each 
other, and I just grabbed the wrong one.


Regardless, upon commenting the correct disk out of fstab, and adding 
back my new secondary swap, everything works again as expected.


It may help others to know what I did.

Part of my original failure to find anything was because I was focused 
on the /lib/systemd, /usr/lib/systemd and /etc/systemd directories.  
Even though I had read the documentation that referenced /run/systemd, I 
had not checked it out.


When I did, I immediately found the correct UUID information for the 
actual failing disk.  And the rest is history.


Thanks for taking the time to look into this.

Bob

Re: Cannot get systemd to forget about swap space on a failed disk

[deloptes]

Bob McGowan wrote:

> Please see thread with subject "Be careful when editing /etc/fstab" for
> a bit of background.
> 
> My computer had two swap partitions, on two different disks, when one of
> them started to generate CRC errors, seek errors, etc.
> 
> Once I determined which of the two it was, I commented out the
> /etc/fstab entry for it, halted the system, removed the faulty disk and
> rebooted.  I got dumped into the "emergency rescue" single user mode,
> after waiting several minutes for various "start jobs" to time out.
> 
> Given the above referenced email, and the info in the links it
> references, I would have thought that rebooting, with the modified fstab
> file, should have worked just fine.
> 
> However, as noted, it did not, so I must assume systemd has a unit (is
> this the right term?) defining my original swap setup.
> 
> I did find in man pages, that if both fstab and a systemd unit define
> the same thing, systemd wins.
> 
> That just reinforces the idea that somewhere, systemd knows the swap
> disk devices.
> 
> But I cannot figure out where this might be, or even if this is the
> correct interpretation.
> 
> Any suggestions on next steps greatly appreciated.
> 
> Bob

If you are using LVM - check LVM. If you keep the disk in the machine -
remove the swap partition. Repeat the test.

# find /lib/systemd/ | grep swap
/lib/systemd/system/swap.target

also

# grep -i swap -r /lib/systemd/
Binary file /lib/systemd/systemd-udevd matches
/lib/systemd/system/sysinit.target:Wants=local-fs.target swap.target
/lib/systemd/system/sysinit.target:After=local-fs.target swap.target
emergency.service emergency.target
/lib/systemd/system/swap.target:Description=Swap
/lib/systemd/system/systemd-journald-dev-log.socket:# Mount and swap units
need this. If this socket unit is removed by an
/lib/systemd/system/systemd-journald-dev-log.socket:# isolate request the
mount and swap units would be removed too,
/lib/systemd/system/systemd-journald.socket:# Mount and swap units need
this. If this socket unit is removed by an
/lib/systemd/system/systemd-journald.socket:# isolate request the mount and
swap units would be removed too,

Re: Cannot get systemd to forget about swap space on a failed disk

[Reco]

Hi.

On Mon, Jun 29, 2020 at 11:10:44PM -0700, Bob McGowan wrote:
> But I cannot figure out where this might be, or even if this is the correct 
> interpretation.

Check out the contents of /etc/systemd/system first.
Rebuild initramfs second.

Reco

Cannot get systemd to forget about swap space on a failed disk

[Bob McGowan]

Please see thread with subject "Be careful when editing /etc/fstab" for 
a bit of background.


My computer had two swap partitions, on two different disks, when one of 
them started to generate CRC errors, seek errors, etc.


Once I determined which of the two it was, I commented out the 
/etc/fstab entry for it, halted the system, removed the faulty disk and 
rebooted.  I got dumped into the "emergency rescue" single user mode, 
after waiting several minutes for various "start jobs" to time out.


Given the above referenced email, and the info in the links it 
references, I would have thought that rebooting, with the modified fstab 
file, should have worked just fine.


However, as noted, it did not, so I must assume systemd has a unit (is 
this the right term?) defining my original swap setup.


I did find in man pages, that if both fstab and a systemd unit define 
the same thing, systemd wins.


That just reinforces the idea that somewhere, systemd knows the swap 
disk devices.


But I cannot figure out where this might be, or even if this is the 
correct interpretation.


Any suggestions on next steps greatly appreciated.

Bob

Re: Swap space choice on a SSD <- Current best practice on?

[Pascal Hambourg]

Le 14/02/2019 à 03:14, Curt Howland a écrit :


I also put in ramdisk options for /tmp in /etc/fstab
You mean tmpfs, not ramdisk. Nobody sane would prefer ramdisk over tmpfs 
for /tmp.

Re: Swap space choice on a SSD <- Current best practice on?

[David Christensen]

On 2/13/19 1:28 PM, Andy Smith wrote:

On Wed, Feb 13, 2019 at 01:14:36PM -0800, David Christensen wrote:

A swap partition is faster than a swap file.


Has something changed in this regard since kernel version 2.6 then?

http://lkml.iu.edu/hypermail/linux/kernel/0507.0/1690.html


I do not follow Linux kernel development.  Apparently, people wanted 
faster swap files badly enough to implement optimizations in kernel 2.6. 
 I can only wonder how much complexity and KLOC are required (overall 
and per file system), if all Linux file systems support it, and are 
there benchmarks?



David

Re: Swap space choice on a SSD <- Current best practice on?

[David Christensen]

On 2/13/19 1:23 PM, Dan Ritter wrote:

David Christensen wrote:

On 2/13/19 6:11 AM, Dan Ritter wrote:

If you want maximum SSD longevity, increase the amount of space that
the SSD can use for remapping by never writing to some amount of
space. Easiest is to not fill the disk with partitions -- leave 5-10%
empty.


AFAIK over-provisioning has no effect on longevity -- longevity is
proportional to total number of cells times rated erase/ write cycles
per cell divided by write throughput.


But, over-provisioning can improve write performance:

https://en.wikipedia.org/wiki/Write_amplification#Over-provisioning


Effective longevity depends on the number of times a block is
erased and written to. You can reduce that over the course of an
SSD's lifetime by limiting the space you think you have
available.

From the article you cited:

"Over-provisioning often takes away from user capacity, either
temporarily or permanently, but it gives back reduced write
amplification, increased endurance, and increased performance."

Increased endurance is increased longevity.


SSD manufacturers taking away part of the capacity and then telling us 
the drive has increased endurance measured against the reduced size is 
marketing speak.  (Similar rant for kB, MB, GB, etc..)



I use the term "over-provisioning" in the context of this mailing list 
-- e.g. what we can do as Debian users (and system administrators):


"Furthermore, if any SSD is set up with an overall partitioning layout 
smaller than 100% of the available space, that unpartitioned space will 
be automatically used by the SSD as over-provisioning as well."



Whether you partition the first 90% of an SSD and write X blocks at 
random intervals over some time period T, or partition 100% and do the 
same, the number of Program/ Erase (P/E) cycles will be the same.  (But 
the timing/ performance of clustered writes may differ.)  When X gets 
large enough, the drive will eventually fail.  I would call that X the 
endurance of the drive.  (Intel converts it to a MTBF of 1,200,000 hours 
for my Series 520 SSD's.)



So, over-provisioning *by Debian users* has no effect on longevity.


David

Re: Swap space choice on a SSD <- Current best practice on?

[Curt Howland]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256


One of the things I do with an SSD is turn down "swappiness" to a 
minimum.

In /etc/sysctl.d/custom.conf

I put the following lines:

vm.swappiness = 0
vm.vfs_cache_pressure = 40
vm.dirty_background_ratio = 10
vm.dirty_ratio = 40

There are also a bunch of networking defaults for IPv4 and v6 for 
security and simplicity, but this is about SSD wear and swap space. A 
swap partition that is simply never written to never wears out.

I also put in ramdisk options for /tmp in /etc/fstab in order to 
reduce disk writes.

Curt-

- -- 
You may my glories and my state dispose,
But not my griefs; still am I king of those.
 --- William Shakespeare, "Richard II"

-BEGIN PGP SIGNATURE-

iHUEAREIAB0WIQTaYVhJsIalt8scIDa2T1fo1pHhqQUCXGTPAgAKCRC2T1fo1pHh
qd6BAP9S86HeF/lA69fSHACF1dPFLRRbM2rKMZUQkcVslWcCRgEAiBXnP41B+hUS
d/01FE0tUlqFFxh7z9axEEjGfOyMv6o=
=visj
-END PGP SIGNATURE-

Re: Swap space choice on a SSD <- Current best practice on?

[Andy Smith]

Hello,

On Wed, Feb 13, 2019 at 04:23:56PM -0500, Dan Ritter wrote:
> "Over-provisioning often takes away from user capacity, either
> temporarily or permanently, but it gives back reduced write
> amplification, increased endurance, and increased performance."
> 
> Increased endurance is increased longevity.

That is also my understanding and matches many articles advising how to
choose the best enterprise SSD for a particular workload. However, I
know that SSDs are a lot more "black box" than your typical HDD so I
think especially with consumer devices it could be hard to generalise
and reason about. At that level the device specs often do not specify
numbers for "terabytes written" or "drive writes per day".

It can also be surprising sometimes how little is written. For example,
I have some servers with flash memory for their operating system
install, with data on other storage:

https://www.supermicro.com/products/nfo/SATADOM.cfm

At the 16GB capacity these offer only 17TB of writes over 5 years and I
was a bit worried, so I was thinking of spending some effort making sure
that things which are regularly doing writes do so to a RAM disk
instead.

Luckily there's a SMART attribute (241) you can use to tell how much has
been written to the drive to date and when I checked that I found the
servers were typically writing only ~14GiB per month. So that would take
about 100 years to reach 17TB! Of course, the 5 year warranty covers
other factors too.

It all depends on use case, as clearly there are uses that are
write-intensive which would burn through 17TB in a matter of hours. I do
not put swap on these devices. Measuring is still essential in my view,
but things are indeed a lot easier than they were a decade ago.

Cheers,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting

Re: Swap space choice on a SSD <- Current best practice on?

[Michael Stone]

On Wed, Feb 13, 2019 at 01:14:36PM -0800, David Christensen wrote:

AFAIK over-provisioning has no effect on longevity -- longevity is
proportional to total number of cells times rated erase/ write cycles
per cell divided by write throughput.


In the absence of trim, restricting the logical capacity of the drive 
ensures a larger pool of cells known to the drive to be unused and 
available for wear leveling. Otherwise, if you write to a cell the drive 
has to always preserve the data in that cell even if you later erase the 
file you wrote there (because the drive doesn't know it was erased).  
This is essentially what the drive manufacturers do: they build a drive 
with capacity N but sell it as N-S where S is the amount of spare 
capacity held in reserve for wear leveling and to allow for cells to be 
removed from service as they reach end of life. In general, the more 
expensive the drive, the more it is overprovisioned to increase the 
longevity of the device.


On modern SSDs in light duty the difference is probably negligeable, 
especially if you occasionally fstrim to communicate how much of the 
drive is actually unused.

Re: Swap space choice on a SSD <- Current best practice on?

[Andy Smith]

Hello,

On Wed, Feb 13, 2019 at 01:14:36PM -0800, David Christensen wrote:
> A swap partition is faster than a swap file.

Has something changed in this regard since kernel version 2.6 then?

http://lkml.iu.edu/hypermail/linux/kernel/0507.0/1690.html

Cheers,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting

Re: Swap space choice on a SSD <- Current best practice on?

[Dan Ritter]

David Christensen wrote: 
> On 2/13/19 6:11 AM, Dan Ritter wrote:
> > If you want maximum SSD longevity, increase the amount of space that
> > the SSD can use for remapping by never writing to some amount of
> > space. Easiest is to not fill the disk with partitions -- leave 5-10%
> > empty.
> 
> AFAIK over-provisioning has no effect on longevity -- longevity is
> proportional to total number of cells times rated erase/ write cycles
> per cell divided by write throughput.
> 
> 
> But, over-provisioning can improve write performance:
> 
> https://en.wikipedia.org/wiki/Write_amplification#Over-provisioning

Effective longevity depends on the number of times a block is
erased and written to. You can reduce that over the course of an
SSD's lifetime by limiting the space you think you have
available.

>From the article you cited:

"Over-provisioning often takes away from user capacity, either
temporarily or permanently, but it gives back reduced write
amplification, increased endurance, and increased performance."

Increased endurance is increased longevity.

-dsr-

Re: Swap space choice on a SSD <- Current best practice on?

[David Christensen]

On 2/13/19 5:41 AM, deb wrote:

Again -- fussing with a full (not from a live .iso) 9.7 install; the
Debian GUI installer is suggesting a Swap partition on a Kingston
SSD.

#1 Given that it's not great to pound the same area of a SSD with 
writes; is it indeed still best practice to go with a swap partition

on a SSD rather than a swap FILE?

(Or is this a legacy spinning hard disk install suggestion?)


On 2/13/19 5:46 AM, Michael Stone wrote:

That's not a thing: the SSD will balance writes physically across
the drive regardless of where they are logically.


+1

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


On 2/13/19 5:41 AM, deb wrote:

#2 How DO you get the installer to go with a Swap FILE?

Just delete that recommended Swap partition during the install?

I looked; but did not run across any best practice docs for Swap on
SSD.


On 2/13/19 6:11 AM, Dan Ritter wrote:

The installer doesn't have that option...


+1 AFAIK


A swap partition is faster than a swap file.


Backing up, archiving, imaging, restoring, etc., swap files is
wasteful, but modifying those operations to exclude/ regenerate swap
files adds complexity (and risk).


In my SOHO environment, all my machines have a single system drive and
my bulk data is in a file server.  When I first started using SSD's and
USB flash drives as system drives, I was worried about swap wearing out
the drive.  So, I tried running without swap.  This worked until
memory got low, then the machines crashed.  So, I added memory and
reinstalled with 1 GB swap partitions.


On 2/13/19 6:11 AM, Dan Ritter wrote:

If you want maximum SSD longevity, increase the amount of space that
the SSD can use for remapping by never writing to some amount of
space. Easiest is to not fill the disk with partitions -- leave 5-10%
empty.


AFAIK over-provisioning has no effect on longevity -- longevity is
proportional to total number of cells times rated erase/ write cycles
per cell divided by write throughput.


But, over-provisioning can improve write performance:

https://en.wikipedia.org/wiki/Write_amplification#Over-provisioning


Other SSD considerations include choice of file system, mount options,
kernel tuning, etc.:

https://btrfs.wiki.kernel.org/index.php/FAQ

https://en.wikipedia.org/wiki/Trim_(computing)

mount(8)

https://cromwell-intl.com/open-source/performance-tuning/disks.html


David

Thanks Dan. Re: Swap space choice on a SSD <- Current best practice on?

[deb]

Thank you.

On 2/13/2019 9:11 AM, Dan Ritter wrote:

deb wrote:

On 2/13/2019 8:46 AM, Michael Stone wrote:

On Wed, Feb 13, 2019 at 08:41:33AM -0500, deb wrote:

#1 Given that it's not great to pound the same area of a SSD with
writes; is it indeed still best practice to go with a swap partition
on a SSD rather than a swap FILE?

That's not a thing: the SSD will balance writes physically across the
drive regardless of where they are logically.



OK, that's interesting.

I though the partition might physically lock a section of writes.

Is there a link that goes into that *partitions are not a boundary on
physical SSD writes* further?

For a spinning disk, your OS generally has control of what will
go where*. For an SSD, the incentive to not rewrite a sector
until absolutely necessary is so high that the geometry of the
disk is completely mythical. The SSD controller will pretend
that there are contiguous sectors for partitions to be in, but
they can all be remapped anywhere at any time.

If you want maximum SSD longevity, increase the amount of space
that the SSD can use for remapping by never writing to some
amount of space. Easiest is to not fill the disk with partitions
-- leave 5-10% empty.

*The disk may remap a known bad sector to a pool of reserve
sectors without the OS being told about it. Or not.


If I'm sure, I'll just let Debian partition for Swap.

... I'm still curious how to get the installer to go with a swap FILE rather
than a swap PARTITION though.

The installer doesn't have that option, but if you tell it to go
ahead without a swap partition, you can create a swapfile
later. Or multiple swapfiles. Or use the swapspace package,
which will create and destroy swapfiles as memory requirements
dictate.

-dsr-

Re: Swap space choice on a SSD <- Current best practice on?

[Dan Ritter]

deb wrote: 
> 
> On 2/13/2019 8:46 AM, Michael Stone wrote:
> > On Wed, Feb 13, 2019 at 08:41:33AM -0500, deb wrote:
> > > #1 Given that it's not great to pound the same area of a SSD with
> > > writes; is it indeed still best practice to go with a swap partition
> > > on a SSD rather than a swap FILE?
> > 
> > That's not a thing: the SSD will balance writes physically across the
> > drive regardless of where they are logically.
> > 
> > 
> OK, that's interesting.
> 
> I though the partition might physically lock a section of writes.
> 
> Is there a link that goes into that *partitions are not a boundary on
> physical SSD writes* further?

For a spinning disk, your OS generally has control of what will
go where*. For an SSD, the incentive to not rewrite a sector
until absolutely necessary is so high that the geometry of the
disk is completely mythical. The SSD controller will pretend
that there are contiguous sectors for partitions to be in, but
they can all be remapped anywhere at any time.

If you want maximum SSD longevity, increase the amount of space
that the SSD can use for remapping by never writing to some
amount of space. Easiest is to not fill the disk with partitions
-- leave 5-10% empty.

*The disk may remap a known bad sector to a pool of reserve
sectors without the OS being told about it. Or not.

> If I'm sure, I'll just let Debian partition for Swap.
> 
> ... I'm still curious how to get the installer to go with a swap FILE rather
> than a swap PARTITION though.

The installer doesn't have that option, but if you tell it to go
ahead without a swap partition, you can create a swapfile
later. Or multiple swapfiles. Or use the swapspace package, 
which will create and destroy swapfiles as memory requirements
dictate. 

-dsr-

Re: Swap space choice on a SSD <- Current best practice on?

[deb]

On 2/13/2019 8:46 AM, Michael Stone wrote:

On Wed, Feb 13, 2019 at 08:41:33AM -0500, deb wrote:
#1 Given that it's not great to pound the same area of a SSD with 
writes; is it indeed still best practice to go with a swap partition 
on a SSD rather than a swap FILE?


That's not a thing: the SSD will balance writes physically across the 
drive regardless of where they are logically.




OK, that's interesting.

I though the partition might physically lock a section of writes.

Is there a link that goes into that *partitions are not a boundary on 
physical SSD writes* further?


If I'm sure, I'll just let Debian partition for Swap.


... I'm still curious how to get the installer to go with a swap FILE 
rather than a swap PARTITION though.



Thanks

Re: Swap space choice on a SSD <- Current best practice on?

[Michael Stone]

On Wed, Feb 13, 2019 at 08:41:33AM -0500, deb wrote:
#1 Given that it's not great to pound the same area of a SSD with 
writes; is it indeed still best practice to go with a swap partition 
on a SSD rather than a swap FILE?


That's not a thing: the SSD will balance writes physically across the 
drive regardless of where they are logically.

Swap space choice on a SSD <- Current best practice on?

[deb]

Hello folks:

Again -- fussing with a full (not from a live .iso) 9.7 install; the 
Debian GUI installer is suggesting a Swap partition on a Kingston SSD.



#1 Given that it's not great to pound the same area of a SSD with 
writes; is it indeed still best practice to go with a swap partition on 
a SSD rather than a swap FILE?


(Or is this a legacy spinning hard disk install suggestion?)


#2 How DO you get the installer to go with a Swap FILE?

Just delete that recommended Swap partition during the install?


I looked; but did not run across any best practice docs for Swap on SSD.


Thank you

Re: Swap space not used

[Lisi]

On Friday 04 May 2012 06:16:52 Bret Busby wrote:
> It could simply be malicious web sites.
>
> I have just tried (repeatedly) to access whitepages.com.au, using
> konqueror (one of the web browsers that I have kept allowing
> Javascript), and, each time that I try to use the web site, it just
> freezes konqueror, requiring me to use the kill switch on konqueror.

It works fine here (Konqueror on Lenny). :-/

Curiouser and curiouser

Perhaps I am being more frugal in my memory use?

Lisi


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/201205041446.16653.lisi.re...@gmail.com

Re: Swap space not used

[Stephen Powell]

On Thu, 03 May 2012 03:48:59 -0400 (EDT), Claudius Hubig wrote:
> 
> Stephen Powell  wrote:
>> It is my understanding that,
>> assuming suspend/resume is supported, your swap partition
>> should be AT LEAST as large as TWICE the amount of RAM.
>> Suspend/resume will consume a RAM's worth right out of the
>> starting gate.  The rest is then available for regular swap
>> file activity.
> 
> This is - more or less - wrong. Suspend/Resume will consume at most
> swap space corresponding to the used RAM (i. e. with compression and
> dropping of buffers/caches, it can be far less). However, this swap
> space is not used during runtime but only on suspend, so if there is
> no need to suspend under heavy load (used swap usually indicates
> heavy load on a desktop and I fail to imagine a reason why you’d like
> to suspend a server…), swap the size of RAM is definitely enough.

Thank you for explaining this.  I was making an assumption based
on previous experience with other operating systems that is
apparently not warranted.  For example, z/VM will allocate its
dump space in the system spool during startup, just to make sure
that if it needs to take a dump, there will be sufficient spool
space available to do it.  The size of the dump space is based on
the amount of RAM, of course.  (Or as mainframers call it, CSTORE.)
Of course, a system dump and a suspend image are taken for different
purposes, but both involve a dump of RAM.

-- 
  .''`. Stephen Powell
 : :'  :
 `. `'`
   `-


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/847935503.954160.1336130435289.javamail.r...@md01.wow.synacor.com

Re: Swap space not used (problem website)

[Scott Ferguson]

On 04/05/12 16:34, Bret Busby wrote:
> On Fri, 4 May 2012, Scott Ferguson wrote:
> 
>>
>> On 04/05/12 15:16, Bret Busby wrote:
>>



>> Works just as well in iceweasel 12.0.1 with NoScript fully enabled.
>> Ditto Konqueror 4.4.5
>>
> 
> Hmm.
> 
> It does work in iceweasel 3.5.16, with Javascript disabled
> 
> ("
> Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.16) Gecko/20120421
> Iceweasel/3.5.16 (like Firefox/3.5.16)
> ")
> 
> I was wrong - I had a memory that it did not work without Javascript.


perhaps you meant to remember "must. not. enable. javascript. for.
whitepages"


:-)

> 
> I have both java and javascript disabled in iceweasel, so it should not
> be that it is using java rather than javascript.

easy to determine (view source).

> 
> But, with Konqueror, when I select the Residential tab, then enter a
> surname, and then an initial, and then tab to the Location or whatever
> text box, it freezes with showing 99% downloaded, and, if I try to close
> the window (after going to alpine, then minimising this window, nothing
> except the title bar is displayed for the Konqueror window), I get
> 
> "
> whitepages.com.au - Search for an Australian Business, Government or
> Person – Konqueror is not responding.
> 
> You may choose to wait a short while for it to continue or force the
> application to quit entirely.
> "
> 
> causing me to select the Force Quit option.
> 
> I have just tried (yet) again, and, in trying to go to the default
> (business) web page, it got to "140 images of 161 downloaded" and 90%,
> then it froze.
> 
> ("
> Konqueror
> Version 4.4.5 (KDE 4.4.5)
> Using KDE Development Platform 4.4.5 (KDE 4.4.5)
> ")

I can't account for your experience - perhaps it's that you're not
running xul-ext-adblock-plus?

> 
> My experience with that web site, is that it is simply malicious.

Which presumes "they did it deliberately", and, implies you've not had
much experience at working for Telstra ;-p



Kind regards

-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa38c50.9060...@gmail.com

Re: Swap space not used

[Bret Busby]

On Fri, 4 May 2012, Scott Ferguson wrote:



On 04/05/12 15:16, Bret Busby wrote:





I have just tried (repeatedly) to access whitepages.com.au, using
konqueror (one of the web browsers that I have kept allowing
Javascript), and, each time that I try to use the web site, it just
freezes konqueror, requiring me to use the kill switch on konqueror.


whitepages.com.au *doesn't* require javascript

links whitepages.com.au
arrow down and select Residential
arrow down and enter Surname
arrow down and enter first intial
arrow down and enter Suburb and State.
hit Enter twice
page down to both your results. (using 1080KB RAM)

Works just as well in iceweasel 12.0.1 with NoScript fully enabled.
Ditto Konqueror 4.4.5



Hmm.

It does work in iceweasel 3.5.16, with Javascript disabled

("
Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.16) Gecko/20120421 
Iceweasel/3.5.16 (like Firefox/3.5.16)

")

I was wrong - I had a memory that it did not work without Javascript.

I have both java and javascript disabled in iceweasel, so it should not 
be that it is using java rather than javascript.


But, with Konqueror, when I select the Residential tab, then enter a 
surname, and then an initial, and then tab to the Location or whatever 
text box, it freezes with showing 99% downloaded, and, if I try to close 
the window (after going to alpine, then minimising this window, nothing 
except the title bar is displayed for the Konqueror window), I get


"
whitepages.com.au - Search for an Australian Business, Government or 
Person – Konqueror is not responding.


You may choose to wait a short while for it to continue or force the 
application to quit entirely.

"

causing me to select the Force Quit option.

I have just tried (yet) again, and, in trying to go to the default 
(business) web page, it got to "140 images of 161 downloaded" and 90%, 
then it froze.


("
Konqueror
Version 4.4.5 (KDE 4.4.5)
Using KDE Development Platform 4.4.5 (KDE 4.4.5)
")

My experience with that web site, is that it is simply malicious.


--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992

Re: Swap space not used (now screen relics).

[Scott Ferguson]

On 04/05/12 15:30, Bret Busby wrote:



>>
> 
> And, Iceweasel (and it may have happened with the iceape browser; I am
> not sure - have not used it for a couple of weeks, now, I think) has a
> habit of leaving fragments of dialogue boxes on top of everything else
> on the desktop, hiding parts of the screen with the dialogue box
> fragments, which only disappear when I kill or crash iceweasel. The only
> thing that displays on top of the dialogue box fragments, is when I
> invoke the screensaver, which overwrites evrything on the screen (at
> least, for the present time)
> 
> I would include a screenshot showing this, but I do not know whether
> this mailing list allows attachments.

I've seen that happen before - I don't believe the problem lies with
Iceweasel though, more likely a video driver/window manager problem




If you gave us (in a fresh post) information about your video etc
someone might be able to help.



Kind regards.


-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa36dba.7060...@gmail.com

Re: Swap space not used

[Scott Ferguson]

On 04/05/12 15:16, Bret Busby wrote:



> 
> I have just tried (repeatedly) to access whitepages.com.au, using
> konqueror (one of the web browsers that I have kept allowing
> Javascript), and, each time that I try to use the web site, it just
> freezes konqueror, requiring me to use the kill switch on konqueror.

whitepages.com.au *doesn't* require javascript

links whitepages.com.au
arrow down and select Residential
arrow down and enter Surname
arrow down and enter first intial
arrow down and enter Suburb and State.
hit Enter twice
page down to both your results. (using 1080KB RAM)

Works just as well in iceweasel 12.0.1 with NoScript fully enabled.
Ditto Konqueror 4.4.5



Not that the site doesn't suck, or that other gubmint sites don't suck
either. :-(



Kind regards

-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa36c63.8000...@gmail.com

Re: Swap space not used

[Bret Busby]

On Fri, 4 May 2012, Scott Ferguson wrote:


On 04/05/12 14:23, Bret Busby wrote:












Perhaps, on installation, the creation of a file to store the original
information about the installation (iso image source, full version
number and date of version, etc), that could be retrieved any time
during the life of the particular installation, would be useful.


/var/log/installer/lsb-release






Thank you for that.

Someone had asked the source of my installation iso.

"
:~$ cat /var/log/installer/lsb-release
DISTRIB_ID=Debian
DISTRIB_DESCRIPTION="Debian GNU/Linux installer"
DISTRIB_RELEASE="6.0 (squeeze) - installer build 20110106+squeeze3+b1"
X_INSTALLATION_MEDIUM=cdrom
"


--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205041336140.32...@bret-dd-workstation.busby.net

Re: Swap space not used

[Bret Busby]

On Fri, 4 May 2012, Scott Ferguson wrote:


On 04/05/12 04:54, Bret Busby wrote:








Out of interest, with you saying that swapping is not mandatory,
from memory, about 20-odd years ago, when I started learning
(formally) about operating systems, we were told that UNIX has a
memory requirement of about 32GB (or, it may have been 32MB - I am
not sure - it was many, many, years ago, and thus, UNIX requires
memory paging to work, and this is why UNIX has had such (relatively)
good memory handling, because it used memory paging; paging out to
hard disk, and, without the paging, UNIX could not have operated.



32MB - back then.
All that is correct, especially then. Now RAM is cheaper and so it's
used in preference to disk (swap).  Some apps, notably
Firefox/Iceweasel, will use all the RAM they can get - and be called
hogs, whereas they're just trying to be efficient and will release
memory when another app calls for it.



And, Iceweasel (and it may have happened with the iceape browser; I am 
not sure - have not used it for a couple of weeks, now, I think) has a 
habit of leaving fragments of dialogue boxes on top of everything else 
on the desktop, hiding parts of the screen with the dialogue box 
fragments, which only disappear when I kill or crash iceweasel. The 
only thing that displays on top of the dialogue box fragments, is when I 
invoke the screensaver, which overwrites evrything on the screen (at 
least, for the present time)


I would include a screenshot showing this, but I do not know whether 
this mailing list allows attachments.



--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205041318230.32...@bret-dd-workstation.busby.net

Re: Swap space not used

[Bret Busby]

On Fri, 4 May 2012, Ralf Mardorf wrote:



Hm, I've got 4 GB RAM and two swaps, 2.17GiB and 2.43GiB, one on each
HDD I'm using.
I'm doing resource-intensive work with my machine.
4 GB RAM are enough for my needs and I never noticed that a swap was
touched.
For my kind of usage Linux (Debian and several other distros) are able
to handle the RAM without fault.

The rule that a swap should be double as large as the RAM is outdated.

Anyway, some people might need much RAM.

Sometime ago I noticed that some (perhaps all, I didn't checked this)
x86_64 kernels on my machine only access 3.8GB from the 4GB RAM on my
machine (no shared memory for the framebuffer), 32-bit PAE kernels are
all ok. I searched the web and found out that other people, having much
more RAM mounted, have this issue with x86_64 kernels too.


The computer has 8GB of RAM, and I have found that the
tendency of web site develpoers, is increased sloppiness, as
too many web site
developers appear to work on the principle that computers have
an
infinite amount of RAM for them to squander.


I don't think so, since there are still a lot of people using 32-bit
Windows installs that can access less than 4GB. If you've got issues
with 8GB RAM when surfing the Internet, I suspect that some script set
limits for the RAM.



It could simply be malicious web sites.

I have just tried (repeatedly) to access whitepages.com.au, using 
konqueror (one of the web browsers that I have kept allowing 
Javascript), and, each time that I try to use the web site, it just 
freezes konqueror, requiring me to use the kill switch on konqueror.


It is mostly Australian government web sites and Australian government 
corporate websites, that use Javacript, that cause most problems, as 
they are simply malicious.


Some of the airlines' web sites are similarly malicious.

--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205041309080.32...@bret-dd-workstation.busby.net

Re: Swap space not used

[Scott Ferguson]

On 04/05/12 14:23, Bret Busby wrote:



> If some utility
> existed that would display the source of an iso image, and the full
> version number of the source iso image, it would be good.

# mount -o loop debian-testing-i386-netinst.iso /mnt
# cat /mnt/.disk/info
Debian GNU/Linux testing "Wheezy" - Official Snapshot i386 NETINST
Binary-1 20120427-21:46

> 
> Perhaps, on installation, the creation of a file to store the original
> information about the installation (iso image source, full version
> number and date of version, etc), that could be retrieved any time
> during the life of the particular installation, would be useful.

/var/log/installer/lsb-release





Kind regards

-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa3666e.8070...@gmail.com

Re: Swap space not used

[Dom]

On 04/05/12 02:28, Ralf Mardorf wrote:

On Fri, 2012-05-04 at 01:36 +0200, Ralf Mardorf wrote:

While this computer has 8GB of RAM, which is far greater than the total
hard drive capacities of most hard drives from twenty years ago


I can't resist ... in the 80s and 90s we burned EPROMS with much less
capacity than an USB stick has got, but they where large ICs, no SMD
technology.

The cartridge of my DX7 synths from the 80s has got a HM6264LP-15 RAM.
For entertainment:
http://www.alldatasheet.com/datasheet-pdf/pdf/HITACHI/HM6264LP-15.html
All inputs and outputs are TTL compatible :D.


Heh! Coincidentally, I was looking through my old chip box last night to 
see if I had a 6502 (I did), and found a HM6116LP-2 and a couple of 
HY62256LP-10 chips :-)


--
Dom


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/4fa364e6.6020...@rpdom.net

Re: Swap space not used

[Bob Proulx]

Rick Thomas wrote:
> Another use for a large swap partition is if you want to put /tmp
> into tmpfs.

Yes.  The new trend for tmpfs /tmp partitions is going to require a
lot of thinking and rethinking for how much swap is required.

Or also swap is useful if you have an enterprise server and have
disabled Linux memory overcommit for reliability.  I have posted about
this problem before.  Here are a couple of them:

  http://lists.debian.org/debian-user/2007/08/msg00022.html

  http://lists.debian.org/debian-user/2008/04/msg02554.html

> Whether doing so is a "good thing(TM)" is a religious debate that I
> don't want to stir up here.  But there are people who do it, and for
> them a large swap partition can be useful.

Agreed.  Very strongly!

Bob


signature.asc
Description: Digital signature

Re: Swap space not used

[Bret Busby]

On Fri, 4 May 2012, Ralf Mardorf wrote:






When I installed it, I had a swap partition of about 40GB set up, as
is shown by gparted.


Did you chose this large swap or was it done automatically? My
installs / + /home have around 20 or 30 GB only. Of cause, for audio
productions I have separated, large partitions.



I chose the swap partition size.

I reasoned that I have (what I regard as) a massively large hard drive 
capacity, so I might as well use it to help system stability, by 
providing a large amount of swap space.



Is your Debian a regular Debian? IOW did you download the image from the
Internet or did you use a DVD from a computer magazine?
At least German computer magazines often have broken versions of distros
as a supplement.



My Debian was installed from downloaded iso images. I think it was from 
Debian 6.03, but I am not sure of the fractional version number 
component. I believe that it was an official Debian version, from the 
debian.org downloads web site, but I am not sure. If some utility 
existed that would display the source of an iso image, and the full 
version number of the source iso image, it would be good.


Perhaps, on installation, the creation of a file to store the original 
information about the installation (iso image source, full version 
number and date of version, etc), that could be retrieved any time 
during the life of the particular installation, would be useful.



--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205041210500.29...@bret-dd-workstation.busby.net

Re: Swap space not used

[Ralf Mardorf]

On Fri, 2012-05-04 at 01:36 +0200, Ralf Mardorf wrote:
> > While this computer has 8GB of RAM, which is far greater than the total 
> > hard drive capacities of most hard drives from twenty years ago

I can't resist ... in the 80s and 90s we burned EPROMS with much less
capacity than an USB stick has got, but they where large ICs, no SMD
technology.

The cartridge of my DX7 synths from the 80s has got a HM6264LP-15 RAM.
For entertainment:
http://www.alldatasheet.com/datasheet-pdf/pdf/HITACHI/HM6264LP-15.html
All inputs and outputs are TTL compatible :D.

Since the battery is empty, I won't change the battery only ;).

Some smart Linux coders where born, when the transistor wasn't invented.
No, I'm not that old :D, I like tubes anyway :).


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1336094900.5833.104.camel@precise

Re: Swap space not used

[Scott Ferguson]

On 04/05/12 04:54, Bret Busby wrote:
> On Thu, 3 May 2012, Scott Ferguson wrote:
> 
>> 
>> On 02/05/12 17:48, Bret Busby wrote:
> 
> 
> 
>> 
>>> 
>>> Why is this so?
>> 
>> JSM is that you? :-)
>> 
> 
> Nope
> 
> :)
> 
> Is he still around?

No (only in spirit). His son is though - and does excellent medical
documentaries. Different hairstyle though.

> 
> 
>> fact there is *no* swap "rule".
>> 
>> Swap is not "required". Enable it if you wish - but it's not
>> mandatory, and it's usefulness is determined by your needs.
>> 
>> For a "desktop" that does a lot of graphic editing you'd normally
>> want
>>> 1GB of RAM and >512MB of swap, more swap than that will usually
>>> result
>> in slower performance. But it will vary considerably from one
>> individual to another. The bigger the pond the more fish you can
>> stock - the smaller the pond the easier it is to catch a given
>> fish.
>> 
>> ie. for netbooks using solid state drives I normally provide *no*
>> swap, if they've 2GB of RAM and don't use suspend (the usual
>> build).
>> 
>> 
>> The system's use of swap is determined by the chosen applications
>> and the "swappiness" settings:- $ cat /proc/sys/vm/swappiness
>> 
>> Default for a "desktop" that's used for development and graphic
>> editing is 60.
>> 
>> 
> 
> I get 60 when I run the command.

That sounds correct.

> 
> Out of interest, with you saying that swapping is not mandatory,
> from memory, about 20-odd years ago, when I started learning
> (formally) about operating systems, we were told that UNIX has a
> memory requirement of about 32GB (or, it may have been 32MB - I am
> not sure - it was many, many, years ago, and thus, UNIX requires
> memory paging to work, and this is why UNIX has had such (relatively)
> good memory handling, because it used memory paging; paging out to
> hard disk, and, without the paging, UNIX could not have operated.


32MB - back then.
All that is correct, especially then. Now RAM is cheaper and so it's
used in preference to disk (swap).  Some apps, notably
Firefox/Iceweasel, will use all the RAM they can get - and be called
hogs, whereas they're just trying to be efficient and will release
memory when another app calls for it.

> 
> Now, while I realise that Linux is not UNIX,

Because no-one wants to pay for the privilege. UNIX is a tag that comes
at a cost. Like Heart Foundation endorsement.

> it is classed, I believe, as "UNIX-like", and som I believe, uses or
> imitates, some of the principles of UNIX.

Yes. Call it a clean room implementation (blindly cloned) that achieves
a high POSIX.

> 
> In that, I believe that Linux requires memory paging, that we rname 
> swapping, and, I understand that the rule used to be to provide swap 
> space of at least twice the amount of RAM.

No. And yes. :-)
It depends very much on how much RAM you have, and what you are running.

For the main desktops with default apps on a system with 1GB of RAM then
I'd suggest 2GB of swap. The same system with 2GB of RAM would probably
only need 1GB of swap (and never fully use it).

The rule I use is total "memory" should be >3GB for a "home desktop".
Aside from that there is no rule - you need to examine the system
requirements (size of files being worked with, disk size and
arrangement, etc) and do some basic testing to determine the optimal
arrangement.

The "swappiness" determines how the system will try and use swap -
testing will show you the best amount of swap. Too much use of swap
slows the system, likewise not enough.

In your case you are barely using swap - and then only a small
percentage of what you have available.


> 
> While this computer has 8GB of RAM, which is far greater than the
> total hard drive capacities of most hard drives from twenty years
> ago, most of the operating systems (including Linux) and the
> applications, have become increasingly bl;oated,

Yes. But - you're using Debian so those excuses don't apply to *you*. ;-p
You *can* turn off the dancing bears and remove the eye-bling. Other
distros leave less choice.

Squeeze KDE will happily run in 512MB of RAM - so that leaves you with
7.5GB of RAM. Dunno about GNOME or the other, lighter desktops.

Modern computers are like modern SUVs - the power requirements are
minimal once you turn off the unnecessary "extras" (airconditioning,
coffee cup heaters, power steering, seat heaters, electric seats and
windows, runnning lights etc.).

> 
> -- Bret Busby Armadale West Australia ..
> 
> "So 

Re: Swap space not used

[Scott Ferguson]

On 04/05/12 00:34, Darac Marjal wrote:
> On Thu, May 03, 2012 at 09:48:59AM +0200, Claudius Hubig wrote:
>> Hello Stephen,
>>
>> Stephen Powell  wrote:
>>> It is my understanding that,
>>> assuming suspend/resume is supported, your swap partition
>>> should be AT LEAST as large as TWICE the amount of RAM.

I seem to recall that was the original suggestion (10 years ago) for
TuxOnIce (usermode suspend/hibernate) - it may even remain true, just
not for kernel mode which seems to work as long as you've slightly more
swapspace (1MB+) than RAM. (I haven't tested with less than that amount).



> 
> If the swap space is available during normal usage, then it's entirely
> possible to have no space to suspend to. 

??

I've *never* found either of the Linux suspend schemes (usermode or
kernel) suffer from that failing - and I used to run a laptop swap
partion only 2MB larger than RAM with Sarge.

Given 15-20 second boot-to-use times on laptops and netbooks, and
session saving, I know longer use hibernate or suspend so I haven't
tested... but I'd be interested in knowing the mechanism behind this
"swap space not available to suspend" phenonema - or references that
document it.

> This is why windows uses a
> separate hibernation file (though Windows' memory management is rather
> poor to start with).

Windoof uses both MS hibernation and OEM hibernation (eg. hidden
hibernate partitions).

I'm not sure why you think Linux hibernation (there's two methods) and
Windoof are related - they have little in common.

> 
> It'd be perfectly reasonable practice to have a separate swap
> file/partition for hibernating to and swapon that before hibernating.

It's perfectly reasonable to have a swap partition on every drive.

You can even swap and/or hibernate to a file if you desire - but a
special swap partition only for hibernation which is "swapon-ed" just
before use sounds unlikely.




Kind regards


-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa31966.80...@gmail.com

Re: Swap space not used

[Ralf Mardorf]

> While this computer has 8GB of RAM, which is far greater than the total 
> hard drive capacities of most hard drives from twenty years ago

40MB (mega bytes!) SCSI drive for my Atari 520 ST here and 4MB RAM (I'm
a tinkerer ;) and it's not only running the Atari TOS, there's a 80286
hardware emulator running DR DOS too.

The C64 in normal usage has got less than 64KB free RAM and already run
GEOS.

But we can't compare those old machines with todays computers, since
today we're doing hardcore multi tasking, render 3D animations etc..


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1336088191.5833.65.camel@precise

Re: Swap space not used

[Ralf Mardorf]

Hm, I've got 4 GB RAM and two swaps, 2.17GiB and 2.43GiB, one on each
HDD I'm using.
I'm doing resource-intensive work with my machine.
4 GB RAM are enough for my needs and I never noticed that a swap was
touched.
For my kind of usage Linux (Debian and several other distros) are able
to handle the RAM without fault.

The rule that a swap should be double as large as the RAM is outdated.

Anyway, some people might need much RAM.

Sometime ago I noticed that some (perhaps all, I didn't checked this)
x86_64 kernels on my machine only access 3.8GB from the 4GB RAM on my
machine (no shared memory for the framebuffer), 32-bit PAE kernels are
all ok. I searched the web and found out that other people, having much
more RAM mounted, have this issue with x86_64 kernels too.

> The computer has 8GB of RAM, and I have found that the
> tendency of web site develpoers, is increased sloppiness, as
> too many web site 
> developers appear to work on the principle that computers have
> an 
> infinite amount of RAM for them to squander.

I don't think so, since there are still a lot of people using 32-bit
Windows installs that can access less than 4GB. If you've got issues
with 8GB RAM when surfing the Internet, I suspect that some script set
limits for the RAM.

> When I installed it, I had a swap partition of about 40GB set up, as
> is shown by gparted.

Did you chose this large swap or was it done automatically? My
installs / + /home have around 20 or 30 GB only. Of cause, for audio
productions I have separated, large partitions.

Is your Debian a regular Debian? IOW did you download the image from the
Internet or did you use a DVD from a computer magazine?
At least German computer magazines often have broken versions of distros
as a supplement.

Regards,
Ralf


2 Cents,
Ralf


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1336087328.5833.57.camel@precise

Re: Swap space not used

[Rick Thomas]

On Fri, 4 May 2012 02:40:16 +0800 (WST), Bret Busby wrote:

free:

"
:~# free
 total   used   free sharedbuffers cached
Mem:   80599647746808 313156  0  54708 
1352976

-/+ buffers/cache:63391241720840
Swap: 42860340  66296   42794044
"
A snaphot of the header of top gives

"
:~# top

top - 02:31:37 up 3 days, 44 min,  3 users,  load average: 0.09, 
0.17, 0.16

Tasks: 205 total,   2 running, 203 sleeping,   0 stopped,   0 zombie
Cpu(s): 24.3%us,  2.6%sy,  0.0%ni, 73.1%id,  0.0%wa,  0.0%hi,  
0.0%si, 0.0%st
Mem:   8059964k total,  7722984k used,   336980k free,55484k 
buffers
Swap: 42860340k total,66296k used, 42794044k free,  1353544k 
cached


"


Well.

According to the above:
You do, indeed, have 40GB of swap configured.
It appears to be using 66MB of that swap.

So swapping is not disabled, and you have plenty of swap space for any 
normal use.


The remaining question is, why do you periodically run out of memory 
and crash?  Or, put another way, what abnormal use is occurring to cause 
your crashes?


Are there any indications in syslog of what may be going on immediately 
before the crash?


Rick


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/6532d84107430b080540568c0c5f0...@pobox.com

Re: Swap space not used

[Bret Busby]

On Thu, 3 May 2012, Scott Ferguson wrote:



On 02/05/12 17:48, Bret Busby wrote:








Why is this so?


JSM is that you?
:-)



Nope

:)

Is he still around?



fact there is *no* swap "rule".

Swap is not "required". Enable it if you wish - but it's not mandatory,
and it's usefulness is determined by your needs.

For a "desktop" that does a lot of graphic editing you'd normally want

1GB of RAM and >512MB of swap, more swap than that will usually result

in slower performance. But it will vary considerably from one individual
to another. The bigger the pond the more fish you can stock - the
smaller the pond the easier it is to catch a given fish.

ie. for netbooks using solid state drives I normally provide *no* swap,
if they've 2GB of RAM and don't use suspend (the usual build).


The system's use of swap is determined by the chosen applications and
the "swappiness" settings:-
$ cat /proc/sys/vm/swappiness

Default for a "desktop" that's used for development and graphic editing
is 60.




I get 60 when I run the command.

Out of interest, with you saying that swapping is not mandatory, from 
memory, about 20-odd years ago, when I started learning (formally) about 
operating systems, we were told that UNIX has a memory requirement of 
about 32GB (or, it may have been 32MB - I am not sure - it was many, 
many, years ago, and thus, UNIX requires memory paging to work, and this 
is why UNIX has had such (relatively) good memory handling, because it 
used memory paging; paging out to hard disk, and, without the paging, 
UNIX could not have operated.


Now, while I realise that Linux is not UNIX, it is classed, I believe, 
as "UNIX-like", and som I believe, uses or imitates, some of the 
principles of UNIX.


In that, I believe that Linux requires memory paging, that we rname 
swapping, and, I understand that the rule used to be to provide swap 
space of at least twice the amount of RAM.


While this computer has 8GB of RAM, which is far greater than the total 
hard drive capacities of most hard drives from twenty years ago, most of 
the operating systems (including Linux) and the applications, have 
become increasingly bl;oated, and the applications become increasingly 
poorly designed, so that memory may not be freed, when an application is 
closed, and, some web sityes that use the dreaded Javascript, keep using 
increasing amounts of RAM, to the extent that I have found it not safe 
to a web browser open, that allows javascvript. I use different web 
browsers for different uses, and the ones that I tend to leave open, are 
Javascript-disabled, for the sake of (relative) system stability.


--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205040241040.21...@bret-dd-workstation.busby.net

Re: Swap space not used

[Bret Busby]

On Wed, 2 May 2012, Andrei POPESCU wrote:


Date: Wed, 2 May 2012 17:27:42
From: Andrei POPESCU 
To: debian-user@lists.debian.org
Subject: Re: Swap space not used

On Mi, 02 mai 12, 15:48:30, Bret Busby wrote:

Hello.

I am running Debian 6.

When I installed it, I had a swap partition of about 40GB set up, as
is shown by gparted.


four zero Gigabytes? My / + /home are only 27GB :)


But, for some strnge reason, Debian 6will not use the swap space,
even though gparted shows it to be "Active".

Instead of Debian 6 using the swap[ partition, it just runs out of
memory, progressively, requiring rebooting every few days.


Please show the output of 'free', 'cat /etc/fstab' and 'fdisk -l' (the
last one will need root).

Kind regards,
Andrei
--



fstab:

"
:~# cat /etc/fstab
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name 
devices

# that works even if disks are added and removed. See fstab(5).
#
#
proc/proc   procdefaults0   0
# / was on /dev/sda6 during installation
UUID=367ac9bc-7790-47b7-aa61-2242b283a9bd /   ext3 
errors=remount-ro 0   1

# /home was on /dev/sda8 during installation
UUID=a3074725-349d-4647-8b07-3a5526f7ee55 /home   ext3 
defaults0   2

# swap was on /dev/sda7 during installation
UUID=dd1aaec4-3b27-4144-a1e3-7b97a75130d3 noneswapsw 
0   0

/dev/scd0   /media/cdrom0   udf,iso9660 user,noauto 0   0
"

free:

"
:~# free
 total   used   free sharedbuffers 
cached
Mem:   80599647746808 313156  0  54708 
1352976

-/+ buffers/cache:63391241720840
Swap: 42860340  66296   42794044
"

fdisk -l :

"
:~# fdisk -l

Disk /dev/sda: 640.1 GB, 640135028736 bytes
255 heads, 63 sectors/track, 77825 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xc000

   Device Boot  Start End  Blocks   Id  System
/dev/sda1   1   9   72261   de  Dell Utility
/dev/sda2  101134 90296327  HPFS/NTFS
/dev/sda31134   11352820826047  HPFS/NTFS
/dev/sda4   11353   77825   5339443425  Extended
/dev/sda5   11353   2173383385351   83  Linux
/dev/sda6   *   21734   3193181915403+  83  Linux
/dev/sda7   31932   3726742860351+  82  Linux swap / 
Solaris

/dev/sda8   42131   5232981923436   83  Linux
/dev/sda9   52330   6252781915403+  83  Linux
/dev/sda10  62528   7272681923436   83  Linux
/dev/sda11  72727   77825409576867  HPFS/NTFS
/dev/sda12  37267   4213039061504   83  Linux

Partition table entries are not in disk order

Disk /dev/sdf: 499.4 GB, 499405291520 bytes
255 heads, 63 sectors/track, 60715 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000f8373

   Device Boot  Start End  Blocks   Id  System
/dev/sdf1   1   60716   4876994567  HPFS/NTFS
"

Someone said something about memtest;

"
:~# memtest
-su: memtest: command not found
"

A snaphot of the header of top gives

"
:~# top

top - 02:31:37 up 3 days, 44 min,  3 users,  load average: 0.09, 0.17, 
0.16

Tasks: 205 total,   2 running, 203 sleeping,   0 stopped,   0 zombie
Cpu(s): 24.3%us,  2.6%sy,  0.0%ni, 73.1%id,  0.0%wa,  0.0%hi,  0.0%si, 
0.0%st

Mem:   8059964k total,  7722984k used,   336980k free,55484k buffers
Swap: 42860340k total,66296k used, 42794044k free,  1353544k cached

"

The computer has Windows 7 Professional, Ubuntu 11.04 and Debian 6 
installed, each in the 64 bit version opf the respective operating 
system, and uses GRUB as the boot selector (I use the term "boot 
selector", as it does not have multiple booting; to me, "multiple 
booting" means booting muliple systems at the same time, and, if I could 
run all of the operating systems at once, I think that I would need 
VMWare or something similar)


The computer has 8GB of RAM, and I have found that the tendency of web 
site develpoers, is increased sloppiness, as too many web site 
developers appear to work on the principle that computers have an 
infinite amount of RAM for them to squander.


When previously running Debian 5 on an AMD K6 (from memory) based 
machine with 2GB of RAM, if the swap was not working, and the RAM usage 
was getting high, if I ran an application such as the GIMP or, I think, 
the Opera web browser, and

Re: Swap space not used

[Claudius Hubig]

Hello Darac,

Darac Marjal  wrote:
> If the swap space is available during normal usage, then it's entirely
> possible to have no space to suspend to.

Yes. However, this is rather unlikely when the computer is used as a
desktop/laptop, don’t you think? The only times when I actually used
my swap space was when some program went rampant and decided to
require more memory than available. During normal operation, my swap
space is seldomly used.

> It'd be perfectly reasonable practice to have a separate swap
> file/partition for hibernating to and swapon that before hibernating.

Certainly, but, for example on a SSD, space is still ‘scarce’ and it
makes therefore sense to think whether one has to be able to suspend
at all costs. If this is still necessary, one could also think about
using only one swap space and swapon it before suspend (and swapoff
afterwards).

Best regards,

Claudius
-- 
"Life is too important to take seriously."
-- Corky Siegel
http://chubig.net  telnet nightfall.org 4242


signature.asc
Description: PGP signature

Re: Swap space not used

[Darac Marjal]

On Thu, May 03, 2012 at 09:48:59AM +0200, Claudius Hubig wrote:
> Hello Stephen,
> 
> Stephen Powell  wrote:
> > It is my understanding that,
> > assuming suspend/resume is supported, your swap partition
> > should be AT LEAST as large as TWICE the amount of RAM.
> > Suspend/resume will consume a RAM's worth right out of the
> > starting gate.  The rest is then available for regular swap
> > file activity.
> 
> This is - more or less - wrong. Suspend/Resume will consume at most
> swap space corresponding to the used RAM (i. e. with compression and
> dropping of buffers/caches, it can be far less). However, this swap
> space is not used during runtime but only on suspend, so if there is

If the swap space is available during normal usage, then it's entirely
possible to have no space to suspend to. This is why windows uses a
separate hibernation file (though Windows' memory management is rather
poor to start with).

It'd be perfectly reasonable practice to have a separate swap
file/partition for hibernating to and swapon that before hibernating.

> no need to suspend under heavy load (used swap usually indicates
> heavy load on a desktop and I fail to imagine a reason why you’d like
> to suspend a server…), swap the size of RAM is definitely enough.
> 
> Best regards,
> 
> Claudius




signature.asc
Description: Digital signature

Re: Swap space not used

[Claudius Hubig]

Hello Stephen,

Stephen Powell  wrote:
> It is my understanding that,
> assuming suspend/resume is supported, your swap partition
> should be AT LEAST as large as TWICE the amount of RAM.
> Suspend/resume will consume a RAM's worth right out of the
> starting gate.  The rest is then available for regular swap
> file activity.

This is - more or less - wrong. Suspend/Resume will consume at most
swap space corresponding to the used RAM (i. e. with compression and
dropping of buffers/caches, it can be far less). However, this swap
space is not used during runtime but only on suspend, so if there is
no need to suspend under heavy load (used swap usually indicates
heavy load on a desktop and I fail to imagine a reason why you’d like
to suspend a server…), swap the size of RAM is definitely enough.

Best regards,

Claudius
-- 
  I marvel at the strength of human weakness.
http://chubig.net  telnet nightfall.org 4242


signature.asc
Description: PGP signature

Re: Swap space not used

[Stephen Powell]

On Wed, 02 May 2012 07:12:31 -0400 (EDT), Sian Mountbatten wrote:
> ...
> As a rule, your swap partition should be the same size as your RAM.
> ...

It is my understanding that,
assuming suspend/resume is supported, your swap partition
should be AT LEAST as large as TWICE the amount of RAM.
Suspend/resume will consume a RAM's worth right out of the
starting gate.  The rest is then available for regular swap
file activity.

-- 
  .''`. Stephen Powell
 : :'  :
 `. `'`
   `-


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/1983883132.927945.1336008398682.javamail.r...@md01.wow.synacor.com

Re: Swap space not used

[Scott Ferguson]

On 02/05/12 17:48, Bret Busby wrote:
> Hello.
> 
> I am running Debian 6.
> 
> When I installed it, I had a swap partition of about 40GB set up, as is
> shown by gparted.
> 
> But, for some strnge reason, Debian 6will not use the swap space,


> even
> though gparted shows it to be "Active".

I don't believe gparted is telling you that swap is "in use".

> 
> Instead of Debian 6 using the swap[ partition, it just runs out of
> memory, progressively, requiring rebooting every few days.

That sounds like a separate problem. Perhaps you could script some tasks
for your system that include a regular "free" command and post the
output to the list so we can analyse this phenonema?

> 
> Why is this so?

JSM is that you?
:-)

> 
> Thank you in anticipation.
> 
> -- 
> Bret Busby
> Armadale
> West Australia
> ..
> 
> "So once you do know what the question actually is,
>  you'll know what the answer means."
> - Deep Thought,
>   Chapter 28 of Book 1 of
>   "The Hitchhiker's Guide to the Galaxy:
>   A Trilogy In Four Parts",
>   written by Douglas Adams,
>   published by Pan Books, 1992
> 
> 
> 
If your system is not using swap then it's most likely been disabled (or
not enabled). Strange, but possible.

As others have suggested - read fstab. Swap should be enabled there.
eg.:-
# grep swap /etc/fstab
# swap was on /dev/sda2 during installation
UUID=999ce39b-f101-417a-9ef8-bf20ec3c86ba noneswapsw
  0   0


Which, if not already enabled would require the following to enable it:-
# swapon /dev/sda2


To see what is currently set for swap:-
# free


There have been a number of comments and "rules" for swap proposed - in
fact there is *no* swap "rule".

Swap is not "required". Enable it if you wish - but it's not mandatory,
and it's usefulness is determined by your needs.

For a "desktop" that does a lot of graphic editing you'd normally want
>1GB of RAM and >512MB of swap, more swap than that will usually result
in slower performance. But it will vary considerably from one individual
to another. The bigger the pond the more fish you can stock - the
smaller the pond the easier it is to catch a given fish.

ie. for netbooks using solid state drives I normally provide *no* swap,
if they've 2GB of RAM and don't use suspend (the usual build).


The system's use of swap is determined by the chosen applications and
the "swappiness" settings:-
$ cat /proc/sys/vm/swappiness

Default for a "desktop" that's used for development and graphic editing
is 60.




Kind regards


-- 
Iceweasel/Firefox/Chrome/Chromium/Iceape/IE extensions for finding
answers to questions about Debian:-
https://addons.mozilla.org/en-US/firefox/collections/Scott_Ferguson/debian/


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4fa1d298.5050...@gmail.com

Re: Swap space not used

[Rick Thomas]

Another use for a large swap partition is if you want to put /tmp into 
tmpfs.


Whether doing so is a "good thing(TM)" is a religious debate that I 
don't want to stir up here.  But there are people who do it, and for 
them a large swap partition can be useful.


Rick

PS:  We haven't heard back from the OP yet.  I'm with the folks who 
think he probably forgot to put it in fstab, but we won't know until he 
replies.



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/39e8e5797ec4782d849f00461fd3f...@pobox.com

Re: Swap space not used

[Shane Johnson]

On Wed, May 2, 2012 at 1:33 PM, Johan Grönqvist
wrote:

> 2012-05-02 13:12, Sian Mountbatten skrev:
>
>  Your swap partition is, very likely, too large. As a rule, your swap
>> partition should be the same size as your RAM. Do you have 40GB RAM?
>>
>
> Linux can handle well above 40 GB of swap. I would be surprised if "swap
> partition too large" was the reason. My swap is larger than that.
>
> I am always a bot surprised by advice like the ones in this thread. I have
> heard and seen this many times.
>
> I am aware that for web-browsing, and other similar activities, using swap
> is almost always bad, as it slows the system down.
>
> I typically run programs and scripts without having a good estimate of
> their future memory usage, and my computer usage is frequently RAM-bound,
> so I try to guess how much I can do within the memory I have available.
>
> It is not uncommon for me to misjudge the need by a factor of 2 or 3, and
> in those cases, I have programs being killed left and right unless I have
> enough swap-space.
>
> I would say that being careful with swap-space is important when one has
> too small a hard drive, but I have plenty of drives space these days. Not
> having my jobs killed is more important to me than saving a few tens of GB
> of extra space.
>
> I have just above 40 GB on my current desktop, and when I had influence
> over a computer with 64GB RAM, it had quite a bit of swap space.
>
> Mem: 64558M total,24822M used,39735M free,  324M buffers
> Swap:   184323M total,   25M used,   184298M free,24238M cached
>
> Regards
>
> Johan
>
>
> --
> To UNSUBSCRIBE, email to 
> debian-user-REQUEST@lists.**debian.orgwith
>  a subject of "unsubscribe". Trouble? Contact
> listmas...@lists.debian.org
> Archive: 
> http://lists.debian.org/**jns26c$n9e$1...@dough.gmane.org<http://lists.debian.org/jns26c$n9e$1...@dough.gmane.org>
>
>
The only time I have seen swap not used when it is active is when it wasn't
needed.  If you can see your swap active using the free command or swapon
-s , and it's not getting used, but you keep crashing, what I usually see
is bad memory.  From my experience memtest would be my next step.

Shane

Re: Swap space not used

[Johan Grönqvist]

2012-05-02 13:12, Sian Mountbatten skrev:


Your swap partition is, very likely, too large. As a rule, your swap
partition should be the same size as your RAM. Do you have 40GB RAM?


Linux can handle well above 40 GB of swap. I would be surprised if "swap 
partition too large" was the reason. My swap is larger than that.


I am always a bot surprised by advice like the ones in this thread. I 
have heard and seen this many times.


I am aware that for web-browsing, and other similar activities, using 
swap is almost always bad, as it slows the system down.


I typically run programs and scripts without having a good estimate of 
their future memory usage, and my computer usage is frequently 
RAM-bound, so I try to guess how much I can do within the memory I have 
available.


It is not uncommon for me to misjudge the need by a factor of 2 or 3, 
and in those cases, I have programs being killed left and right unless I 
have enough swap-space.


I would say that being careful with swap-space is important when one has 
too small a hard drive, but I have plenty of drives space these days. 
Not having my jobs killed is more important to me than saving a few tens 
of GB of extra space.


I have just above 40 GB on my current desktop, and when I had influence 
over a computer with 64GB RAM, it had quite a bit of swap space.


Mem: 64558M total,24822M used,39735M free,  324M buffers
Swap:   184323M total,   25M used,   184298M free,24238M cached

Regards

Johan


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/jns26c$n9e$1...@dough.gmane.org

Re: Swap space not used

[Claudius Hubig]

Hello Lisi,

Lisi  wrote:
> On Wednesday 02 May 2012 12:12:31 Sian Mountbatten wrote:
> > As a rule, your swap
> > partition should be the same size as your RAM.
> 
> We used to be taught it should be twice as big as your RAM - but even that 
> wouldn't get you to 40GB!!  And, of course, that was in the days when RAM was 
> tiny by today's standards.

That indeed was a rule of thumb when swap space was actually
important. However, with today’s RAM, the main usage of swap space is
hibernate (suspend to disk), for which at most the size of your RAM
in swap space is required.

Best regards,

Claudius
-- 
Remember, UNIX spelled backwards is XINU.
-- Mt.
http://chubig.nettelnet nightfall.org 4242


signature.asc
Description: PGP signature

Re: Swap space not used

[Lisi]

On Wednesday 02 May 2012 12:12:31 Sian Mountbatten wrote:
> As a rule, your swap
> partition should be the same size as your RAM.

We used to be taught it should be twice as big as your RAM - but even that 
wouldn't get you to 40GB!!  And, of course, that was in the days when RAM was 
tiny by today's standards.

Lisi


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/201205021447.17851.lisi.re...@gmail.com

Re: Swap space not used

[Sian Mountbatten]

On 02/05/12 09:00, Bret Busby wrote:

Hello.

I am running Debian 6.

When I installed it, I had a swap partition of about 40GB set up, as is
shown by gparted.

But, for some strnge reason, Debian 6will not use the swap space, even
though gparted shows it to be "Active".

Instead of Debian 6 using the swap[ partition, it just runs out of
memory, progressively, requiring rebooting every few days.

Why is this so?

Thank you in anticipation.

Your swap partition is, very likely, too large. As a rule, your swap
partition should be the same size as your RAM. Do you have 40GB RAM?

Sincerely

--
Sian Mountbatten
ex-Algol 68 specialist


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/jnr4qu$hag$1...@speranza.aioe.org

Re: Swap space not used

[Andrei POPESCU]

On Mi, 02 mai 12, 15:48:30, Bret Busby wrote:
> Hello.
> 
> I am running Debian 6.
> 
> When I installed it, I had a swap partition of about 40GB set up, as
> is shown by gparted.
 
four zero Gigabytes? My / + /home are only 27GB :)

> But, for some strnge reason, Debian 6will not use the swap space,
> even though gparted shows it to be "Active".
> 
> Instead of Debian 6 using the swap[ partition, it just runs out of
> memory, progressively, requiring rebooting every few days.

Please show the output of 'free', 'cat /etc/fstab' and 'fdisk -l' (the 
last one will need root).

Kind regards,
Andrei
-- 
Offtopic discussions among Debian users and developers:
http://lists.alioth.debian.org/mailman/listinfo/d-community-offtopic


signature.asc
Description: Digital signature

Re: Swap space not used

[keith]

On Wed, 2 May 2012 15:48:30 +0800 (WST)
Bret Busby  wrote:

> Hello.
> 
> I am running Debian 6.
> 
> When I installed it, I had a swap partition of about 40GB set up, as is 
> shown by gparted.
> 
> But, for some strnge reason, Debian 6will not use the swap space, even 
> though gparted shows it to be "Active".
> 
> Instead of Debian 6 using the swap[ partition, it just runs out of 
> memory, progressively, requiring rebooting every few days.
> 
> Why is this so?


Perhaps it isn't in /etc/fstab.

You could try using swapon / swapoff on the partition

-- 
keith 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/20120502093150.7e4cd90ec8b61634fd294...@gmail.com

Re: Swap space not used

[Andy Hawkins]

Hi,

In article ,
   Bret Busby wrote:
> When I installed it, I had a swap partition of about 40GB set up, as is 
> shown by gparted.
>
> But, for some strnge reason, Debian 6will not use the swap space, even 
> though gparted shows it to be "Active".
>
> Instead of Debian 6 using the swap[ partition, it just runs out of 
> memory, progressively, requiring rebooting every few days.
>
> Why is this so?

Is it listed in fstab?

I have:

/dev/sda2   noneswapsw  0   0

Andy


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/slrnjq1r0r.84u.a...@atom.gently.org.uk

Swap space not used

[Bret Busby]

Hello.

I am running Debian 6.

When I installed it, I had a swap partition of about 40GB set up, as is 
shown by gparted.


But, for some strnge reason, Debian 6will not use the swap space, even 
though gparted shows it to be "Active".


Instead of Debian 6 using the swap[ partition, it just runs out of 
memory, progressively, requiring rebooting every few days.


Why is this so?

Thank you in anticipation.

--
Bret Busby
Armadale
West Australia
..

"So once you do know what the question actually is,
 you'll know what the answer means."
- Deep Thought,
  Chapter 28 of Book 1 of
  "The Hitchhiker's Guide to the Galaxy:
  A Trilogy In Four Parts",
  written by Douglas Adams,
  published by Pan Books, 1992



--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: 
http://lists.debian.org/alpine.deb.2.00.1205021543070.14...@bret-dd-workstation.busby.net

Re: [Slightly OT] Releasing swap space?

[Brian]

On Sun 14 Aug 2011 at 15:37:34 +0100, AG wrote:

> My hard ram is 3 GB (2.84 to be exact) and I gave the same amount to  
> swap when I initially partitioned the HDD.

More than enough.

> My response to Ivan crossed yours, so if there's no value and I also run  
> the risk of meddling with the kernel's affairs, it seems wise to leave  
> well enough alone, unless I have a specific need to do so and the usage  
> has accumulated.

   brian@desktop:~$ uptime
19:51:39 up 88 days, 21:29,  6 users,  load average: 0.10, 0.17, 0.22

   brian@desktop:~$ free
total   used   free sharedbuffers cached
   Mem:   10358281000468  35360  0  67996 343620
   -/+ buffers/cache: 588852 446976
   Swap:  1951888  501321901756

I have no idea why 50M of swap is used but the machine chugs along. I'd
suggest you forget about it.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110814190250.GT14528@desktop

Re: [Slightly OT] Releasing swap space?

[Javier Vasquez]

On Sun, Aug 14, 2011 at 6:51 AM, AG  wrote:
> Hey list
>
> Just a quick query about releasing swap space.  On occasion according to
> Conky (system monitoring app), the swap space (set at 3Gb) sometimes gets
> used to up to 15% especially if using something like Pan for usenet.
>
> Is there any value/ harm in releasing this space using something like:
>
> swapoff -a && swapon -a [1]
>
> Thanks for your opinions.
>
> AG
>
>
> [1] found on the web, but not AFAIK Debian specific


If your intend is to try minimizing swapping, trying to keep in RAM as
much as possible, why not using the swappiness kernel configuration?

See:

% 'grep' swap /etc/sysctl.conf
# Start looking at using swap only when ram has that % left.
vm.swappiness=6

That means, attempt to start using swap only when there's 6% remaining
free RAM.  Now, it's NOT guaranteed that's what the kernel will do,
it's just a hint.  You can look on the web for swappiness, you'll find
all kind of blogs and comments...

That might help minimize the use of swap area...


-- 
Javier.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/calurrgfknpntpcsb7fthdbpanjpckkrueyx6_wuwste4obk...@mail.gmail.com

Re: [Slightly OT] Releasing swap space?

[Ivan Shmakov]

> Camaleón   writes:

[…]

 > So, who is going to say that a "/swap" partition is going to be
 > needed with 8 GiB of RAM?  I wouldn't, I just thought kernel makes
 > use of all of the available resources are allocates them to get the
 > best performance.  Meaning: if you have available resources (i.e.,
 > unused swap) they will be used.

When preparing some files to be written to a DVD+R, I'd usually
put them to /tmp/.  The total volume of the files in such a case
may easily exceed the amount of the physical RAM I have on the
host, so the swap gets used.

IOW, I consider swap to be not only the disk space that gets
used in the case there's insufficient physical RAM, but also the
space that I can easily use for temporary files.

Therefore, it's my preference, and everlasting recommendation,
to have about 3% of the disk space allocated for swap.

-- 
FSF associate member #7257


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/86hb5k56mw@gray.siamics.net

Re: [Slightly OT] Releasing swap space?

[Ivan Shmakov]

> AG   writes:
> On 14/08/11 14:35, Ivan Shmakov wrote:
> AG   writes:

[…]

 > I have recently switched to Xfce4 on Stable from Gnome because the
 > latter was quite a memory hog and seemed to retain pages in swap
 > until I logged/ rebooted.

The pages that stay longer in the swap are the pages that are
used less frequently.

 > Xfce4 fortunately doesn't do this, but I did notice that after using
 > Swiftfox I was finding that 3% usage of swap didn't clear when I
 > closed the app.

These are likely to be the pages belonging to some “sleeping”
processes.

 > This set me to wondering how much control I could exert and even
 > should exert over swap's content.

 > Are you able to confirm whether the code given previously is accurate
 > *and* safe should I want to pursue clearing swap on the fly?

The code in question is only safe if there's enough free RAM to
load all the swapped pages into.

Otherwise, OOM madness may ensue.

-- 
FSF associate member #7257


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/86liuw5744@gray.siamics.net

Re: [Slightly OT] Releasing swap space?

[Camaleón]

On Sun, 14 Aug 2011 15:37:34 +0100, AG wrote:

> On 14/08/11 15:27, Camaleón wrote:
 
>> What's your amount of physical ram?
>>
>>
> My hard ram is 3 GB (2.84 to be exact) and I gave the same amount to
> swap when I initially partitioned the HDD.

That's a fair amount of ram... I wonder why your system is in the need of 
making use of swap.

>>> Is there any value/ harm in releasing this space using something like:
>>>
>>> swapoff -a&&  swapon -a [1]
>> (...)
>>
>> No harm, but no need to do it neither, unless you have a specific
>> requirement. Swap usage is up to the kernel, just let it to manage as
>> it desires (remember the kernel's law: "unused memory is wasted
>> memory").
> 
> I wasn't aware of that law.  Thanks for the info ;-)
> 
> My response to Ivan crossed yours, so if there's no value and I also run
> the risk of meddling with the kernel's affairs, it seems wise to leave
> well enough alone, unless I have a specific need to do so and the usage
> has accumulated.

I would investigate why 3 GiB is not enough, maybe there is a background 
app that is (ab)using too much ram for any reason :-?

It just happened to me something similar a couple of days ago, but on a 
server that has 8 GiB of RAM (and 1 GiB of swap). I was copying a big 
file (~37 GiB) over the network from a windows client to the samba server 
and when it finished, I realized the server was using a small amount of 
swap (¿?) and kept it so until the next day when the server was started 
again :-)

So, who is going to say that a "/swap" partition is going to be needed 
with 8 GiB of RAM? I wouldn't, I just thought kernel makes use of all of 
the available resources are allocates them to get the best performance. 
Meaning: if you have available resources (i.e., unused swap) they will be 
used.

Greetings,

-- 
Camaleón


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/pan.2011.08.14.15.13...@gmail.com

Re: [Slightly OT] Releasing swap space?

[AG]

On 14/08/11 15:27, Camaleón wrote:

On Sun, 14 Aug 2011 13:51:07 +0100, AG wrote:


Just a quick query about releasing swap space.  On occasion according to
Conky (system monitoring app), the swap space (set at 3Gb) sometimes
gets used to up to 15% especially if using something like Pan for
usenet.


Hi Camaleón

Good to hear from you.


What's your amount of physical ram?



My hard ram is 3 GB (2.84 to be exact) and I gave the same amount to 
swap when I initially partitioned the HDD.

Is there any value/ harm in releasing this space using something like:

swapoff -a&&  swapon -a [1]

(...)

No harm, but no need to do it neither, unless you have a specific
requirement. Swap usage is up to the kernel, just let it to manage as it
desires (remember the kernel's law: "unused memory is wasted memory").


I wasn't aware of that law.  Thanks for the info ;-)

My response to Ivan crossed yours, so if there's no value and I also run 
the risk of meddling with the kernel's affairs, it seems wise to leave 
well enough alone, unless I have a specific need to do so and the usage 
has accumulated.


Cheers

AG

Greetings,




--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/4e47ddae.5060...@gmail.com

Re: [Slightly OT] Releasing swap space?

[AG]

On 14/08/11 14:35, Ivan Shmakov wrote:

AG  writes:

  >  Just a quick query about releasing swap space.  On occasion according to
  >  Conky (system monitoring app), the swap space (set at 3Gb) sometimes
  >  gets used to up to 15% especially if using something like Pan for
  >  usenet.

  >  Is there any value/ harm in releasing this space using something like:

  >  swapoff -a&&  swapon -a [1]

Without the context, I see neither harm nor value in doing that.

However, if there was some task which used a lot of virtual
memory, thus forcing the other tasks' pages to be moved to swap,
doing that will bring those pages back to RAM, which may improve
the future responsiveness of the system.  (Up to the time when
such a memory-hungry task is run again; say, a Web browser with
a few dozens of graphically-rich pages in the tabs.)

  >  Thanks for your opinions.

  >  [1] found on the web, but not AFAIK Debian specific



Thank you for that, Ivan.

I have recently switched to Xfce4 on Stable from Gnome because the 
latter was quite a memory hog and seemed to retain pages in swap until I 
logged/ rebooted.  Xfce4 fortunately doesn't do this, but I did notice 
that after using Swiftfox I was finding that 3% usage of swap didn't 
clear when I closed the app.  This set me to wondering how much control 
I could exert and even should exert over swap's content.


Are you able to confirm whether the code given previously is accurate 
*and* safe should I want to pursue clearing swap on the fly?


Cheers

AG


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/4e47dc93.10...@gmail.com

Re: [Slightly OT] Releasing swap space?

[Camaleón]

On Sun, 14 Aug 2011 13:51:07 +0100, AG wrote:

> Just a quick query about releasing swap space.  On occasion according to
> Conky (system monitoring app), the swap space (set at 3Gb) sometimes
> gets used to up to 15% especially if using something like Pan for
> usenet.

What's your amount of physical ram?
 
> Is there any value/ harm in releasing this space using something like:
> 
> swapoff -a && swapon -a [1]

(...)

No harm, but no need to do it neither, unless you have a specific 
requirement. Swap usage is up to the kernel, just let it to manage as it 
desires (remember the kernel's law: "unused memory is wasted memory").

Greetings,

-- 
Camaleón


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/pan.2011.08.14.14.27...@gmail.com

Re: [Slightly OT] Releasing swap space?

[Ivan Shmakov]

>>>>> AG   writes:

 > Just a quick query about releasing swap space.  On occasion according to
 > Conky (system monitoring app), the swap space (set at 3Gb) sometimes
 > gets used to up to 15% especially if using something like Pan for
 > usenet.

 > Is there any value/ harm in releasing this space using something like:

 > swapoff -a && swapon -a [1]

Without the context, I see neither harm nor value in doing that.

However, if there was some task which used a lot of virtual
memory, thus forcing the other tasks' pages to be moved to swap,
doing that will bring those pages back to RAM, which may improve
the future responsiveness of the system.  (Up to the time when
such a memory-hungry task is run again; say, a Web browser with
a few dozens of graphically-rich pages in the tabs.)

 > Thanks for your opinions.

 > [1] found on the web, but not AFAIK Debian specific

-- 
FSF associate member #7257


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/861uwo6rpc@gray.siamics.net

[Slightly OT] Releasing swap space?

[AG]

Hey list

Just a quick query about releasing swap space.  On occasion according to 
Conky (system monitoring app), the swap space (set at 3Gb) sometimes 
gets used to up to 15% especially if using something like Pan for usenet.


Is there any value/ harm in releasing this space using something like:

swapoff -a && swapon -a [1]

Thanks for your opinions.

AG


[1] found on the web, but not AFAIK Debian specific


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: http://lists.debian.org/4e47c4bb.8080...@gmail.com

Re: Use of Swap Space

[Brad Sawatzky]

On Tue, 03 Feb 2009, Aneurin Price wrote:

> On Tue, Feb 3, 2009 at 12:13 PM, Ron Johnson  wrote:
> > On 02/03/2009 05:01 AM, Avi Greenbury wrote:
> >
> > That's a design issue (I think) specific to Windows.  Has to do with  the
> > decision to map video card RAM into regular address space (even on machines
> > with discreet video cards).

This situation is true for most any OS.  Linux (for example) can work
around it fairly well by properly implementing PAE to extend the address
space.  Windows doesn't work so well with PAE (thanks to the joy of closed
source drivers).

> You should be able to avoid this problem by enabling PAE - search Google
> and there'a a lot of information about this, but a lot of it sadly is
> either inconsistent or just clearly wrong (it's my experience that that's
> the case for most responses to questions relating to Windows, but then
> maybe that's just my prejudice showing).
>
> Try looking at 
> http://blogs.msdn.com/oldnewthing/archive/2006/08/14/699521.aspx
> and searching Raymond Chen's blog for more recent posts on the topic.

PAE (usually) doesn't work well under Windows unless all the drivers are
64-bit aware.  Apparently WinXP sp2 just silently ignores the /PAE flag
(at least as far as memory addressing is concerned).

Dan's Data has a good summary of the 3GB+ issues with 32 bit windows here:
  

> Apologies for pushing this thread further OT.

Likewise.

-- Brad


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Aneurin Price]

On Tue, Feb 3, 2009 at 12:13 PM, Ron Johnson  wrote:
> On 02/03/2009 05:01 AM, Avi Greenbury wrote:
>>
>> Ron Johnson wrote:
>>  > On 02/01/2009 10:04 AM, Mirko Scurk wrote:
>>  > [snip]
>>  >>
>>  >> Could it be that 32-bit Debian can't access rest of memory?
>>  >>
>>  >
>>  > That would only be an issue if he could only see (I think) 2GB of his
>>  > 4GB RAM.
>>  >
>>
>> Really?
>> The only system on which I've >3Gb of ram and a 32bit OS is my Windows
>> laptop, and that reports 3.5Gb of the ~4Gb that's in there. I've no idea how
>> to get a less-rounded number for it out of Windows.
>
> That's a design issue (I think) specific to Windows.  Has to do with  the
> decision to map video card RAM into regular address space (even on machines
> with discreet video cards).
>

You should be able to avoid this problem by enabling PAE - search Google and
there'a a lot of information about this, but a lot of it sadly is either
inconsistent or just clearly wrong (it's my experience that that's the case
for most responses to questions relating to Windows, but then maybe that's
just my prejudice showing).

Try looking at http://blogs.msdn.com/oldnewthing/archive/2006/08/14/699521.aspx
and searching Raymond Chen's blog for more recent posts on the topic.
Apologies for pushing this thread further OT.

Nye


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 02/03/2009 05:01 AM, Avi Greenbury wrote:

Ron Johnson wrote:
 > On 02/01/2009 10:04 AM, Mirko Scurk wrote:
 > [snip]
 >>
 >> Could it be that 32-bit Debian can't access rest of memory?
 >>
 >
 > That would only be an issue if he could only see (I think) 2GB of his
 > 4GB RAM.
 >

Really?
The only system on which I've >3Gb of ram and a 32bit OS is my Windows 
laptop, and that reports 3.5Gb of the ~4Gb that's in there. I've no idea 
how to get a less-rounded number for it out of Windows.


That's a design issue (I think) specific to Windows.  Has to do with 
 the decision to map video card RAM into regular address space 
(even on machines with discreet video cards).


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Avi Greenbury]

Ron Johnson wrote:
> On 02/01/2009 10:04 AM, Mirko Scurk wrote:
> [snip]
>>
>> Could it be that 32-bit Debian can't access rest of memory?
>>
>
> That would only be an issue if he could only see (I think) 2GB of his
> 4GB RAM.
>

Really?
The only system on which I've >3Gb of ram and a 32bit OS is my Windows 
laptop, and that reports 3.5Gb of the ~4Gb that's in there. I've no idea 
how to get a less-rounded number for it out of Windows.


--
Avi Greenbury
http://aviswebsite.co.uk ;)
http://aviswebsite.co.uk/asking-questions


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 02/02/2009 11:34 AM, Douglas A. Tutty wrote:

On Fri, Jan 30, 2009 at 10:04:19PM -0600, Ron Johnson wrote:

On 01/30/2009 09:54 PM, Douglas A. Tutty wrote:

On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:

On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:



If a 100% CPU usage causes your computer to *over*heat (which is
something you did not write above) you'll have a problem elsewhere (e.g.
a dist-upgrade that includes tex-live :-) )

Or, is the board firmware slowing down the CPU to allow it to cool,
which makes a basically idling debian system to use a higher percentage
of available CPU (less idle time).  Remember, its a percentage of CPU
capacity.  If the CPU slows down, the the available capacity decreases.
But "100% CPU usage" isn't the same as "running at 100% of rated 
clock speed".


Is *that* what you are really referring to?


To illustrate what I mean, lets change from percentage to some
ficticuous numbers.  Lets say that the CPU when cool can make 100
BIGMACS per second.  When the CPU is cool, top shows that it is 99% idle
(1% CPU usage), so it takes 1 BIGMACS per second to keep the system and
top running.  Now, heat up the CPU too far and the BIOS slows down the
CPU to make it run cooler.  Now the CPU can only make 1 BIGMACS per
second.  It still takes 1 BIGMACS per second to keep the system and top
running, so now top will show that the CPU is 0% idle (100% CPU usage).

This is what I was suggesting.


No need for analogies.

$ cat /proc/cpuinfo | grep 'cpu MHz'
cpu MHz : 1000.000
cpu MHz : 1000.000

[Do something CPU-intensive.]

$ cat /proc/cpuinfo | grep 'cpu MHz'
cpu MHz : 2100.000
cpu MHz : 2100.000

Again, I ask if you are referring to clock speed or CPU usage?

--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Stefan Monnier]

>> Try to run `free' to get a more detailed break up (or even "cat
>> /proc/meminfo").

> Running free -g on my system returns 3. I have 4.
> Running cat /proc/meminfo returns 4030668KB, which is 3.84GB according
> to onlineconversion, closer. Does RAM also have a sort of FAT?

No, it's just that some part of your RAM is used up for other things.
E.g. it may be used by your video card (if it doesn't have its own
VRAM), or by the kernel.

> It seems as though free won't return the accurate size.

The opposite: it returns the actual usable size (of course "-g" makes
it round *down* to the nearest gigabyte, so you get a gross
under-approximation, but that's what you asked for).


Stefan


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Douglas A. Tutty]

On Fri, Jan 30, 2009 at 10:04:19PM -0600, Ron Johnson wrote:
> On 01/30/2009 09:54 PM, Douglas A. Tutty wrote:
> >On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:
> >>On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:

> >>If a 100% CPU usage causes your computer to *over*heat (which is
> >>something you did not write above) you'll have a problem elsewhere (e.g.
> >>a dist-upgrade that includes tex-live :-) )
> >
> >Or, is the board firmware slowing down the CPU to allow it to cool,
> >which makes a basically idling debian system to use a higher percentage
> >of available CPU (less idle time).  Remember, its a percentage of CPU
> >capacity.  If the CPU slows down, the the available capacity decreases.
> 
> But "100% CPU usage" isn't the same as "running at 100% of rated 
> clock speed".
> 
> Is *that* what you are really referring to?

To illustrate what I mean, lets change from percentage to some
ficticuous numbers.  Lets say that the CPU when cool can make 100
BIGMACS per second.  When the CPU is cool, top shows that it is 99% idle
(1% CPU usage), so it takes 1 BIGMACS per second to keep the system and
top running.  Now, heat up the CPU too far and the BIOS slows down the
CPU to make it run cooler.  Now the CPU can only make 1 BIGMACS per
second.  It still takes 1 BIGMACS per second to keep the system and top
running, so now top will show that the CPU is 0% idle (100% CPU usage).

This is what I was suggesting.

Doug.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 02/01/2009 08:15 PM, Lee Glidewell wrote:
[snip]


You know what they say about people who assume. ;)

I'll go stand in the corner now. 


With your nose pressed into the corner, touching that dust spot at 
eye level.


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Lee Glidewell]

On Sunday 01 February 2009 17:59:07 Ron Johnson wrote:
> On 02/01/2009 02:49 PM, Lee Glidewell wrote:
> >
> > No, the issue is that manufactures advertise in *1000, while computers
> > use
>
> Hard drive manufacturers, not RAM manufacturers.
>
> My beard's grey enough to remember when drive manufacturers measured
> drive capacity in binary KB, not decimal.
>
> > bytes in *1024. The recent convention that's come into place to represent
> > this is between Kilo/Mega/Giga-bytes (*1000) and Kibi/Mebi/Gibi-bytes
> > (*1024).
> >
> > So a stick of memory advertised as 4 Gigabytes is going to present itself
> > to your computer as 3.84 Gibibytes, roughly.
>
> If that were true, I'd have 8 * 10^9 bytes of RAM, and this
> demonstrates that error:
> 

Okay, I stand corrected. I guess I had assumed that RAM was sold this way as 
well, and hadn't bothered to do the math to check it.

You know what they say about people who assume. ;)

I'll go stand in the corner now. 

Lee


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 02/01/2009 02:49 PM, Lee Glidewell wrote:

On Saturday 31 January 2009 21:01:14 David Fox wrote:

It isn't that RAM has a FAT - those things only are present on
filesystems. It is more likely that free's interpretation doesn't
include kernel memory. Also, 4gb may be 4*1024*1024 not 4*1000*1000,
although that is more likely to be a concern with hard disk capacity.


No, the issue is that manufactures advertise in *1000, while computers use 


Hard drive manufacturers, not RAM manufacturers.

My beard's grey enough to remember when drive manufacturers measured 
drive capacity in binary KB, not decimal.


bytes in *1024. The recent convention that's come into place to represent this 
is between Kilo/Mega/Giga-bytes (*1000) and Kibi/Mebi/Gibi-bytes (*1024). 

So a stick of memory advertised as 4 Gigabytes is going to present itself to 
your computer as 3.84 Gibibytes, roughly. 


If that were true, I'd have 8 * 10^9 bytes of RAM, and this 
demonstrates that error:


$ calc 8 \* 10\*\*9 / 1024
7812500

$ cat /proc/meminfo | grep MemTot
MemTotal:8177796 kB

$ calc 8177796 \* 1024
8374063104

$ calc 2\*\*30 \* 8
8589934592

The only issue is that I'm "missing" 215871488 bytes (52703 of 4KB 
pages).



http://xkcd.com/394/


But seriously: KB was always the standard way of writing kilobyte. 
Kb was/is kilobit (now, Mb and Gb are relevant), used my memory 
manufactures to indicate the capacity of individual *chips* (not the 
SIMMs on which they are mounted).


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Lee Glidewell]

On Sunday 01 February 2009 17:04:38 Nuno Magalhães wrote:
> > So a stick of memory advertised as 4 Gigabytes is going to present itself
> > to your computer as 3.84 Gibibytes, roughly.
>
> Er... what's the standard in Debian? 1024, right? We're still being
> logical here, right?

Sorry, it's more like 3.72 Gibibytes.

Anyway, it has nothing to do with Debian. It has to do with the fact that 
computers are binary, and memory manufacturers advertise their products by 
measuring in multiples of 10. Some might say that practice is dishonest, but 
it is also universal at this point. 

And the Kibibyte/Mebibyte/Gibibyte nomenclature was introduced by the IEEE 
like 10 years ago, so people really should be familiar with it by now. :)




signature.asc
Description: This is a digitally signed message part.

Re: Use of Swap Space

[Nuno Magalhães]

> So a stick of memory advertised as 4 Gigabytes is going to present itself to
> your computer as 3.84 Gibibytes, roughly.

Er... what's the standard in Debian? 1024, right? We're still being
logical here, right?


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Lee Glidewell]

On Saturday 31 January 2009 21:01:14 David Fox wrote:
>
> It isn't that RAM has a FAT - those things only are present on
> filesystems. It is more likely that free's interpretation doesn't
> include kernel memory. Also, 4gb may be 4*1024*1024 not 4*1000*1000,
> although that is more likely to be a concern with hard disk capacity.

No, the issue is that manufactures advertise in *1000, while computers use 
bytes in *1024. The recent convention that's come into place to represent this 
is between Kilo/Mega/Giga-bytes (*1000) and Kibi/Mebi/Gibi-bytes (*1024). 

So a stick of memory advertised as 4 Gigabytes is going to present itself to 
your computer as 3.84 Gibibytes, roughly. 

http://xkcd.com/394/

Lee


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 02/01/2009 10:04 AM, Mirko Scurk wrote:
[snip]


Could it be that 32-bit Debian can't access rest of memory?



That would only be an issue if he could only see (I think) 2GB of 
his 4GB RAM.


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Mirko Scurk]

Ron Johnson wrote:
> On 01/31/2009 03:27 PM, Nuno MagalhĂŁes wrote:
>>> Try to run `free' to get a more detailed break up (or even "cat
>>> /proc/meminfo").
>>
>> Running free -g on my system returns 3. I have 4.
>> Running cat /proc/meminfo returns 4030668KB, which is 3.84GB according
>
> That's 4030668*1024 = 4,127,404,032.
>
>> to onlineconversion, closer. Does RAM also have a sort of FAT?
>
> 4GB = 4,294,967,296.
>
> So, where's the missing 167,563,264 bytes?  (I.e. 40,909 4KB pages.)
>
>> It seems as though free won't return the accurate size.
>
> Of course it does.  "We", though, aren't kernel hackers or h/w
> gurus, so are left in the dark.
>
> --
> Ron Johnson, Jr.
> Jefferson LA  USA
>

Could it be that 32-bit Debian can't access rest of memory?

-- 
Mirko Scurk


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 01/31/2009 11:01 PM, David Fox wrote:
[snip]


It isn't that RAM has a FAT - those things only are present on
filesystems. It is more likely that free's interpretation doesn't
include kernel memory. Also, 4gb may be 4*1024*1024 not 4*1000*1000,
although that is more likely to be a concern with hard disk capacity.


No, RAM is always measured in binary.

--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 01/31/2009 03:27 PM, Nuno Magalhães wrote:

Try to run `free' to get a more detailed break up (or even "cat
/proc/meminfo").


Running free -g on my system returns 3. I have 4.
Running cat /proc/meminfo returns 4030668KB, which is 3.84GB according


That's 4030668*1024 = 4,127,404,032.


to onlineconversion, closer. Does RAM also have a sort of FAT?


4GB = 4,294,967,296.

So, where's the missing 167,563,264 bytes?  (I.e. 40,909 4KB pages.)


It seems as though free won't return the accurate size.


Of course it does.  "We", though, aren't kernel hackers or h/w 
gurus, so are left in the dark.


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[David Fox]

On Sat, Jan 31, 2009 at 1:27 PM, Nuno Magalhães  wrote:

> It seems as though free won't return the accurate size.

I also have 4 gb of RAM (new Quadcore Intel) and 'free -g' reports '3'
as well, I suspect this is underrounding to the extreme, and 'free
-gb' returns a more realistic number:

f...@newbox:~$ free -gb
 total   used   free sharedbuffers cached
Mem:4018106368 3988586496   29519872  0   26746880 2479407104
-/+ buffers/cache: 1482432512 2535673856
Swap:   83897753605529600 8384245760

It isn't that RAM has a FAT - those things only are present on
filesystems. It is more likely that free's interpretation doesn't
include kernel memory. Also, 4gb may be 4*1024*1024 not 4*1000*1000,
although that is more likely to be a concern with hard disk capacity.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Vincent Lefevre]

On 2009-01-31 02:52:34 +, Tzafrir Cohen wrote:
> On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:
> > I recently noticed that my CPU is at 100% when the temperature of
> > my CPU is high. Taking a look at System Monitor i have also
> > discovered that my swap space is being used while only 13% of the
> > RAM is, why isn't is using the rest of the RAM. Has anyone got any
> > ideas why?
> 
> No. CPU utilization has nothing directly to do with swap usage.

If the machine swaps a lot, the load average can get very high.
For instance, on my PowerBook, just because I started a process
that required much memory, the load average reached 25. I suppose
the cause is that background processes that woke up were waiting
for memory, and during this time they were regarded as running.

-- 
Vincent Lefèvre  - Web: <http://www.vinc17.org/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon)


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Ron Johnson]

On 01/31/2009 06:00 PM, Andrew Sackville-West wrote:

On Fri, Jan 30, 2009 at 10:04:19PM -0600, Ron Johnson wrote:

On 01/30/2009 09:54 PM, Douglas A. Tutty wrote:

On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:

On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:

Hi
I recently noticed that my CPU is at 100% when the temperature of my CPU is
high. 

...

If a 100% CPU usage causes your computer to *over*heat (which is
something you did not write above) you'll have a problem elsewhere (e.g.
a dist-upgrade that includes tex-live :-) )

Or, is the board firmware slowing down the CPU to allow it to cool,
which makes a basically idling debian system to use a higher percentage
of available CPU (less idle time).

...
But "100% CPU usage" isn't the same as "running at 100% of rated clock 
speed".



...

regardless of which of the above is happening, if the thing is running
hot, then get in there and clean out the fans and heat
sinks there's nothing like trying to slow down a cpu to make it
run cooler when the cooler isn't working...


I prop up the rear of the laptop.  Now because of a docking station, 
before with a short section of 2x4 wood.  Allows for better air 
flow, and less fan usage.


--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Andrew Sackville-West]

On Fri, Jan 30, 2009 at 10:04:19PM -0600, Ron Johnson wrote:
> On 01/30/2009 09:54 PM, Douglas A. Tutty wrote:
>> On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:
>>> On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:
 Hi
 I recently noticed that my CPU is at 100% when the temperature of my CPU is
 high. 
...
>>> If a 100% CPU usage causes your computer to *over*heat (which is
>>> something you did not write above) you'll have a problem elsewhere (e.g.
>>> a dist-upgrade that includes tex-live :-) )
>>
>> Or, is the board firmware slowing down the CPU to allow it to cool,
>> which makes a basically idling debian system to use a higher percentage
>> of available CPU (less idle time).
...
> But "100% CPU usage" isn't the same as "running at 100% of rated clock 
> speed".
>
...

regardless of which of the above is happening, if the thing is running
hot, then get in there and clean out the fans and heat
sinks there's nothing like trying to slow down a cpu to make it
run cooler when the cooler isn't working...

.02

A


signature.asc
Description: Digital signature

Re: Use of Swap Space

[Nuno Magalhães]

> Try to run `free' to get a more detailed break up (or even "cat
> /proc/meminfo").

Running free -g on my system returns 3. I have 4.
Running cat /proc/meminfo returns 4030668KB, which is 3.84GB according
to onlineconversion, closer. Does RAM also have a sort of FAT?

It seems as though free won't return the accurate size.

Nuno Magalhães


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Stefan Monnier]

> I recently noticed that my CPU is at 100% when the temperature of my CPU is
> high. Taking a look at System Monitor i have also discovered that my swap
> space is being used while only 13% of the RAM is, why isn't is using the
> rest of the RAM. Has anyone got any ideas why?

Unless you really have an enormous amount of RAM (in which case you
probably know what's going on), the tool that tells you "13% of RAM
used" is partly lying.  It's probably telling you that 13% of the RAM is
used by data that's not found anywhere else, while most of the rest of
your RAM is probably used to store local copies of disk blocks.
Try to run `free' to get a more detailed break up (or even "cat
/proc/meminfo").


Stefan


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Jochen Schulz]

Dean Chester:
>
> I recently noticed that my CPU is at 100% when the temperature of my CPU is
> high.

The causality is the other way round: your CPU's temperature rises if
the CPU is being used. That's totally expected. The question is whether
the temperature is high enough to damage your CPU.

J.
-- 
If I was Mark Chapman I would have shot John Lennon with a water pistol.
[Agree]   [Disagree]
 


signature.asc
Description: Digital signature

Re: Use of Swap Space

[Ron Johnson]

On 01/30/2009 09:54 PM, Douglas A. Tutty wrote:

On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:

On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:

Hi
I recently noticed that my CPU is at 100% when the temperature of my CPU is
high. Taking a look at System Monitor i have also discovered that my swap
space is being used while only 13% of the RAM is, why isn't is using the
rest of the RAM. Has anyone got any ideas why?

No. CPU utilization has nothing directly to do with swap usage.

If a 100% CPU usage causes your computer to *over*heat (which is
something you did not write above) you'll have a problem elsewhere (e.g.
a dist-upgrade that includes tex-live :-) )


Or, is the board firmware slowing down the CPU to allow it to cool,
which makes a basically idling debian system to use a higher percentage
of available CPU (less idle time).  Remember, its a percentage of CPU
capacity.  If the CPU slows down, the the available capacity decreases.


But "100% CPU usage" isn't the same as "running at 100% of rated 
clock speed".


Is *that* what you are really referring to?

--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Douglas A. Tutty]

On Sat, Jan 31, 2009 at 02:52:34AM +, Tzafrir Cohen wrote:
> On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:
> > Hi
> > I recently noticed that my CPU is at 100% when the temperature of my CPU is
> > high. Taking a look at System Monitor i have also discovered that my swap
> > space is being used while only 13% of the RAM is, why isn't is using the
> > rest of the RAM. Has anyone got any ideas why?
> 
> No. CPU utilization has nothing directly to do with swap usage.
> 
> If a 100% CPU usage causes your computer to *over*heat (which is
> something you did not write above) you'll have a problem elsewhere (e.g.
> a dist-upgrade that includes tex-live :-) )

Or, is the board firmware slowing down the CPU to allow it to cool,
which makes a basically idling debian system to use a higher percentage
of available CPU (less idle time).  Remember, its a percentage of CPU
capacity.  If the CPU slows down, the the available capacity decreases.

Doug.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Tzafrir Cohen]

On Sat, Jan 31, 2009 at 12:54:34AM +, Dean Chester wrote:
> Hi
> I recently noticed that my CPU is at 100% when the temperature of my CPU is
> high. Taking a look at System Monitor i have also discovered that my swap
> space is being used while only 13% of the RAM is, why isn't is using the
> rest of the RAM. Has anyone got any ideas why?

No. CPU utilization has nothing directly to do with swap usage.

If a 100% CPU usage causes your computer to *over*heat (which is
something you did not write above) you'll have a problem elsewhere (e.g.
a dist-upgrade that includes tex-live :-) )

-- 
Tzafrir Cohen | tzaf...@jabber.org | VIM is
http://tzafrir.org.il || a Mutt's
tzaf...@cohens.org.il ||  best
ICQ# 16849754 || friend


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Use of Swap Space

[Boyd Stephen Smith Jr.]

On Friday 2009 January 30 18:54:34 Dean Chester wrote:
>I recently noticed that my CPU is at 100% when the temperature of my CPU is
>high. Taking a look at System Monitor i have also discovered that my swap
>space is being used while only 13% of the RAM is, why isn't is using the
>rest of the RAM. Has anyone got any ideas why?
>Thanks in Advance
>Dean.

Once stuff gets swapped out, it may not get swapped in.  (It will remain 
swapped out until it is accessed.  Swapping has a cost; leaving it there does 
not.)  Even if it is, it will remain backed by swap until it is written to.  
(Until it is written to, the existing swap space may save a disk write if the 
page needs to be swapped in the future.  Invalidating the page costs a 
little; leaving it there does not.)

Basically use of swap and non-use of RAM are *not* problems.
-- 
Boyd Stephen Smith Jr. ,= ,-_-. =. 
b...@iguanasuicide.net ((_/)o o(\_))
ICQ: 514984 YM/AIM: DaTwinkDaddy   `-'(. .)`-' 
http://iguanasuicide.net/  \_/ 


signature.asc
Description: This is a digitally signed message part.

Re: Use of Swap Space

[Ron Johnson]

On 01/30/2009 06:54 PM, Dean Chester wrote:

Hi
I recently noticed that my CPU is at 100% when the temperature of my CPU 
is high. Taking a look at System Monitor i have also discovered that my 
swap space is being used


How much?  2%, or 90%?

 while only 13% of the RAM is, why isn't is 
using the rest of the RAM. Has anyone got any ideas why?


What does top(1) say?

--
Ron Johnson, Jr.
Jefferson LA  USA

"I am not surprised, for we live long and are celebrated poopers."


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Use of Swap Space

[Dean Chester]

Hi
I recently noticed that my CPU is at 100% when the temperature of my CPU is
high. Taking a look at System Monitor i have also discovered that my swap
space is being used while only 13% of the RAM is, why isn't is using the
rest of the RAM. Has anyone got any ideas why?
Thanks in Advance
Dean.

Re: swap space on a large system

[Douglas A. Tutty]

On Wed, Jun 11, 2008 at 06:51:08AM -0400, Mag Gam wrote:
> Typically, we create a partition to capture a kernel dump when the system
> crashes. Therefore, a system with 16GB of RAM will have a partition with
> 16GB.
> 
> How would I scale a system with 64 or 128GB of memory? Any thoughts?
> 

Bigger hard drive?

Doug.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: swap space on a large system

[Luke S Crawford]

"Mag Gam" <[EMAIL PROTECTED]> writes:

> Typically, we create a partition to capture a kernel dump when the system
> crashes. Therefore, a system with 16GB of RAM will have a partition with
> 16GB.

> How would I scale a system with 64 or 128GB of memory? Any thoughts?

As far as I understand, though, kdump doesn't require you to have a massive
swapfile;  it can be configured to dump other places (though personally,
my swap is always at least my ram size for other reasons)

http://lse.sourceforge.net/kdump/

most linux folks, though, don't seem to be that into crash dumps;  I know
when I made the jump from FreeBSD to Linux, this was a big shocker for me,
as most of the *BSD shops I've been in, policy was that when something crashed,
you didn't trust that box until you figured out why it crashed (usually
by GDBing the crash dump and/or banging on the hardware)

but the linux folks seem to think that the crash dumps are not so useful.
http://lkml.org/lkml/2007/5/25/267  -  and this shows in the support for
crash dumps.  -  they seem to change the mechanism for collecting system
dumps every year or two.At most of the linux shops I've worked,
it's hard enough to get people to get me serial consoles (which we
then set up to log, so we at least can tell the difference between
the power getting pulled and a MCE)  crash dumps are right out.  and I 
can kinda see why;  it's something of a pain in the ass under linux.  

Thing is, it was actually useful under *BSD.  

But that's my rant, and unquestionably, Mr. Torvalds knows a lot more than
I do, so maybe I should just shut up.  


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: swap space on a large system

[Ron Johnson]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On 06/11/08 05:51, Mag Gam wrote:
> Typically, we create a partition to capture a kernel dump when the
> system crashes.

How often does that happen?

> Therefore, a system with 16GB of RAM will have a
> partition with 16GB.
> 
> How would I scale a system with 64 or 128GB of memory? Any thoughts?

- --
Ron Johnson, Jr.
Jefferson LA  USA

"Kittens give Morbo gas.  In lighter news, the city of New New
York is doomed."
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIT88OS9HxQb37XmcRAogZAKDchpZaZjzgwLvLeeDc1ofFbMw8agCfUh2C
ZtozLVj3dRjC69HSeunuZNk=
=eIw6
-END PGP SIGNATURE-


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

swap space on a large system

[Mag Gam]

Typically, we create a partition to capture a kernel dump when the system
crashes. Therefore, a system with 16GB of RAM will have a partition with
16GB.

How would I scale a system with 64 or 128GB of memory? Any thoughts?


TIA

swap space and hd partitioning

[tyler]

Hi,

I've got a couple of questions regarding hd partitions and swap space.

My first, immediate problem is that I've just upgraded my RAM from 1.5GB
to 3GB. I'm running some numerical simulations and analysis that require
that much space or more. My swap partition is 1.95GB, and I've
discovered that when I've got lots of simulation data in memory I don't
have enough space to hibernate.

Reading the Debian Reference, I see the advice to keep to 2GB or less
for each swap partition, but to have at least 1x RAM. In my case, where
I'm using a lot of RAM and I want to be able to hibernate, can I just
set up a second 2GB swap partition, or should I make a single large swap
partition of 3-4GB? I'm not sure how hibernate and swapping work with
multiple swap partitions.

The other, slightly less pressing question, is that I've been dual
booting up to now, and so have quite a fragmented harddrive. In order to
make enough space for the extra swap I'm going to have to further cut
into my windows partition, to the point that I might as well just blow
it away (haven't booted into it in months). Is there a way to reorganize
my file system more elegant than simply backing everything up and
reinstalling the OS? The current layout, as reported by gparted:

/dev/sda1 - 14GB Windows
/dev/sda4 - 41GB extended
  /dev/sda5 - 1.95GB swap
  /dev/sda8 - 7GB /home/tyler/photos/
  /dev/sda6 - 26GB /home
  /dev/sda7 - 6GB /

I'm not sure why the partitions are out of order, but the last
alteration I did, increasing the extended and a few contained partitions
at the expense of windows and the pre-installed recovery partition,
managed to re-order all the partitions, such that fstab and menu.lst
couldn't find anything until I reinstalled grub from a live cd.

My current fstab and menu.lst are pasted below, just for completeness.
Thanks for any suggestions!

Tyler


# /etc/fstab: static file system information.
#
#
proc/proc   procdefaults0   0
/dev/sda7   /   ext3defaults,errors=remount-ro 0   1
/dev/sda6   /home   ext3defaults0   2
/dev/sda8   /home/tyler/photos ext3defaults0   2
/dev/sda5   noneswapsw  0   0
/dev/hda/media/cdrom0   udf,iso9660 user,noauto 0   0

/dev/sdb1   /home/tyler/jumpdrive vfat noauto,users,exec,umask=000 0 0
/dev/sda1   /mnt/windowsntfs noauto,ro,users,noexec,umask=000 0 0


timeout 5
color cyan/blue white/blue
foreground ff
background 0639a1

title   Debian GNU/Linux, kernel 2.6.24-1-686
root(hd0,6)
kernel  /boot/vmlinuz-2.6.24-1-686 root=/dev/sda7 ro 
initrd  /boot/initrd.img-2.6.24-1-686

title   Debian GNU/Linux, kernel 2.6.24-1-686 (single-user mode)
root(hd0,6)
kernel  /boot/vmlinuz-2.6.24-1-686 root=/dev/sda7 ro single
initrd  /boot/initrd.img-2.6.24-1-686

title   Debian GNU/Linux, kernel 2.6.22-3-686
root(hd0,6)
kernel  /boot/vmlinuz-2.6.22-3-686 root=/dev/sda7 ro 
initrd  /boot/initrd.img-2.6.22-3-686
savedefault

title   Debian GNU/Linux, kernel 2.6.22-3-686 (single-user mode)
root(hd0,6)
kernel  /boot/vmlinuz-2.6.22-3-686 root=/dev/sda7 ro single
initrd  /boot/initrd.img-2.6.22-3-686
savedefault

title Microsoft Windows XP Professional at sda1
rootnoverify (hd0,0)
chainloader +1

title MEMTEST
kernel /boot/memtest86+.bin

### BEGIN AUTOMAGIC KERNELS LIST
## lines between the AUTOMAGIC KERNELS LIST markers will be modified
## by the debian update-grub script except for the default options below

## DO NOT UNCOMMENT THEM, Just edit them to your needs

## ## Start Default Options ##
## default kernel options
## default kernel options for automagic boot options
## If you want special options for specific kernels use kopt_x_y_z
## where x.y.z is kernel version. Minor versions can be omitted.
## e.g. kopt=root=/dev/hda1 ro
##  kopt_2_6_8=root=/dev/hdc1 ro
##  kopt_2_6_8_2_686=root=/dev/hdc2 ro
# kopt=root=/dev/sda7 ro

## default grub root device
## e.g. groot=(hd0,0)
# groot=(hd0,6)

## should update-grub create alternative automagic boot options
## e.g. alternative=true
##  alternative=false
# alternative=true

## should update-grub lock alternative automagic boot options
## e.g. lockalternative=true
##  lockalternative=false
# lockalternative=false

## additional options to use with the default boot option, but not with the
## alternatives
## e.g. defoptions=vga=791 resume=/dev/hda5
# defoptions=

## should update-grub lock old automagic boot options
## e.g. lockold=false
##  lockold=true
# lockold=false

## Xen hypervisor options to use with the default Xen boot option
# xenhopt=

## Xen Linux kernel options to use with the default Xen boot option
# xenkopt=console=tty0

## altop