Re: Progress on my new Debian box

[Bzzzz]

On Mon, 23 Jun 2014 19:21:41 -0700
Joseph Loo  wrote:
 
> I think you are missing the problem associated with SSd. The wear 
> problem is associated with the amount of free space. If the drive
> is 99.99% full, you could probably wear the drive out in no time
> at all. The wear problem is prevented by using free space that has
> not been written to it. thus if 00 % full drive will wear out
> faster than a drive that is 1% full. If you do not speak in
> context of the percentage full the benchmarks are not too useful.

Not at all.

What I'm saying is SSD may be good for kitchen usage, not for
heavy use (and especially big files often fully rewritten, as
in CAD systems).

And this is what you confirm below.
 
> Most consumer grade ssd are limited to about 10K writes per cell.
> If you exceed the limit, dead cell. Remember that another factor
> involve is the number of spare cells. all of these things play in
> the role when an ssd fails.

Re-reading the beginning of your §, I'm now asking myself a
terrible question (as the 10K barrier last for long now):
is the 10K a real physical barrier to SSD life, or is it
the same like the 1,000 hours duration for light bulbs.

I know there's a "cage destruction" effect when writing a
SSD cell, nevertheless I wouldn't be surprised that this 
barrier's already broken but stays covered (imagine a 
technological breakthrough that would allow 1000K writes/cell,
it would be spinning HDz' end right'o and a more than
a significant SSD sales decrease rapidly).

I'm not writing a novel, I know a bit of physics and some
people that are researchers in the branch of new/improved
techniques of substrates masking/doping…

-- 
 Fed up with live
 Ur right, Ur 16, Ur future's before U, no money problems, a PC,
 food, a web access, a hot bedroom, a family, Christmas gifts,
 may be pocket money, only some homework to do when back to home,
 boring teachers, friends, a plasma TV, a Wii, but life is not 
 fair, I understand U…
 I don't have a Wii yet
 Damn U…


signature.asc
Description: PGP signature

Re: Progress on my new Debian box

[Joseph Loo]

On 06/23/2014 12:27 PM, B wrote:

On Tue, 24 Jun 2014 04:34:00 +1000
Andrew McGlashan  wrote:


I believe that would be true of quite /old/ SSD drives, but
definitely not for newer ones.


I wouldn't be so positive… until a real independent lab,
conducting real tests (especially with a high number of
small files, test curiously (much too) often absent from
"testers" sites).


The new drives are subject to write issues,


Yeah, like older ones.


but to hit that problem will take just as long as a
traditional spinning drive -- they too have limits, spinning
drives are mechanical.


May be, but most of my disks have a ≥ 10 years life (24/7) with
a very few errors (only 2 of 45 have 1 & 3 unrecoverable sectors),
so, if you can prove me SSD is as good as these, why not…


There have been very heavy torture tests on thew newer range of
SSDs and they are performing exceptionally well with mega data
being written [1], up to fairly heavy data usage levels.


All are biased ("strangely", to lower the write errors due to
multiple write repetitions on the same sectors); this is why
until a _real_ lab, with plausible tests protocols and
methodology doesn't make a test, I won't trust it more than
my first underwear :)


There is apparently a way to restore SSD drives to original
condition by super heating the layer that breaks down (due to
writes), targeting the exact spot with the right temperature
returns the SSD drive to brand new state.  Not sure when this
newest generation will hit the market though. [2]


Yeah, go figure heating _some_ cells among all in a today's
chip density; not to mention that I don't see other
sites/labs/researchers saying the same thing.

On this ground too, us firms can't be trusted as they hire
and pay indelicate specialists to _get_ the result they
_want_; just as monsanto or the govts does.



I think you are missing the problem associated with SSd. The wear 
problem is associated with the amount of free space. If the drive is 
99.99% full, you could probably wear the drive out in no time at all. 
The wear problem is prevented by using free space that has not been 
written to it. thus if 00 % full drive will wear out faster than a drive 
that is 1% full. If you do not speak in context of the percentage full 
the benchmarks are not too useful.


Most consumer grade ssd are limited to about 10K writes per cell. If you 
exceed the limit, dead cell. Remember that another factor involve is the 
number of spare cells. all of these things play in the role when an ssd 
fails.


--
Joseph Loo
j...@acm.org


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

Archive: https://lists.debian.org/53a8e0b5.5040...@acm.org

Re: raid/mdadm help

[Rodney D. Myers]

On Mon, 23 Jun 2014 13:05:08 -0600
Bob Proulx  wrote:

> Rodney D. Myers wrote:
> > Linux-Fan wrote:  
> > > Rodney D. Myers wrote:  
> > > > Is there a debian specific mail list or online forum to get some
> > > > help for a newbie setting up a raid storage device?
> > > > 
> > > > I can find plenty of tips to get started, but things are not
> > > > doing the exact thing(s) the online guides are showing.  
> 
> It is all about using Debian so this mailing list is good.  Lots of us
> are using raid.
> 
> > > Also, if you manually want to create a RAID 1 of two partitions
> > > use
> > > 
> > >   # mdadm --create /dev/md0 --level=1 --raid-devices=2 \
> > >   /dev/sda1 /dev/sdb1  
> > 
> > Not attempting to install on a raid, but create a raid for storage  
> 
> Linux-Fan's suggestion is a good one.  If you are not using it for the
> system then it is easier to get experience building raid yourself as
> in the above example.  François's suggestions were good too.

okay I have a 4 drive bay, and I did this (similar to the above)

 /sbin/mdadm --create --verbose /dev/md0 --level=5
 --raid-devices=4 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1 mdadm: layout
 defaults to left-symmetric mdadm: layout defaults to left-symmetric
mdadm: chunk size defaults to 512K
mdadm: /dev/sdb1 appears to contain an ext2fs file system
   size=976760832K  mtime=Wed Dec 31 19:00:00 1969
mdadm: /dev/sdc1 appears to contain an ext2fs file system
   size=976762580K  mtime=Wed Dec 31 19:00:00 1969
mdadm: /dev/sdc1 appears to be part of a raid array:
   level=raid5 devices=3 ctime=Mon Jun 23 06:55:12 2014
mdadm: /dev/sdd1 appears to contain an ext2fs file system
   size=976760832K  mtime=Mon Jun 23 16:54:04 2014
mdadm: /dev/sde1 appears to contain an ext2fs file system
   size=488385560K  mtime=Wed Dec 31 19:00:00 1969
mdadm: /dev/sde1 appears to be part of a raid array:
   level=raid5 devices=3 ctime=Mon Jun 23 06:55:12 2014
mdadm: size set to 488254464K
mdadm: automatically enabling write-intent bitmap on large array
mdadm: largest drive (/dev/sdc1) exceeds size (488254464K) by more than
 1% Continue creating array? yes
mdadm: Defaulting to version 1.2 metadata
mdadm: array /dev/md0 started.


I let it run for a few hours and when I returned home I did the
following and found it stopped;

/sbin/mdadm --detail /dev/md0
/dev/md0:
Version : 1.2
  Creation Time : Mon Jun 23 17:00:23 2014
 Raid Level : raid5
 Array Size : 1464763392 (1396.91 GiB 1499.92 GB)
  Used Dev Size : 488254464 (465.64 GiB 499.97 GB)
   Raid Devices : 4
  Total Devices : 4
Persistence : Superblock is persistent

  Intent Bitmap : Internal

Update Time : Mon Jun 23 20:09:45 2014
  State : active, FAILED 
 Active Devices : 0
 Failed Devices : 4
  Spare Devices : 0

 Layout : left-symmetric
 Chunk Size : 512K

Number   Major   Minor   RaidDevice State
   0   000  removed
   2   002  removed
   4   004  removed
   6   006  removed

   0   8   17-  faulty   /dev/sdb1
   1   8   33-  faulty   /dev/sdc1
   2   8   49-  faulty   /dev/sdd1
   4   8   65-  faulty   /dev/sde1


Not sure what's going on

-- 
Rodney D. Myers 

They that can give up essential liberty to obtain a
little temporary safety deserve neither liberty nor safety.
        Ben Franklin - 1759


signature.asc
Description: PGP signature

Re: Progress on my new Debian box

[Bzzzz]

On Mon, 23 Jun 2014 16:59:04 -0400
Steve Litt  wrote:

> 3 years later I bought an IBM Deskstar drive with 6 times the
> capacity for $300.00 :-)

These aren't even SATA but IDE; but they're on old machines
that give satisfaction for what they're used: storing CAD 
drawings updated very often and are backed up when needed.
So I don't really need extra space (except if I become very
lazy and let the HDz get bloated:)

In production world, you avoid touching what's working right
until it really breaks (remember that PCI bus superseded the
ISA bus only a fistful of years ago).

I understand your way (especially the HD's one AT THE TIME YOU
DESCRIBE), but today I don't wanna fall in consumerism.
For large calculations, I use big juicy multi-core/CPU machines
with several TB of RAM, but for my every day reporting or stuffs
like that, a 12 years machine mono-CPU & 1.5GB RAM is far enough…

-- 
 What did you do for woman's day?
 I left her out


signature.asc
Description: PGP signature

Re: Progress on my new Debian box

[Steve Litt]

On Mon, 23 Jun 2014 20:05:07 +0100
Brian  wrote:

> On Mon 23 Jun 2014 at 12:43:19 -0400, Steve Litt wrote:
> 
> > Anyway, things are going well. My other observation is that Debian
> > is less like Ubuntu than it initially looks.
> 
> Debian looks less like Ubuntu than you originally thought? What did
> you expect and why should it matter? - they are two different OSs.
> Related maybe; but peas and beans have a lot in common.

I expected it to look like Ubuntu but without the layer of "we do it
all for you" tools, without obnoxiating Plymouth, and more configurable
with Vim.

I didn't expect a lot of the little differences in networking, in
Dovecot, etc. That being said, they're much more similar to each other
than either is to, let's say, Fedora or FreeBSD. Debian and Ubuntu both
use apt-get and dpkg and synaptic and aptitude: If I know one I can
kind of limp along in the other.

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623170453.3c7f5ad0@mydesk

Re: Progress on my new Debian box

[Steve Litt]

On Mon, 23 Jun 2014 21:27:59 +0200
B  wrote:

> May be, but most of my disks have a ≥ 10 years life (24/7) with
> a very few errors (only 2 of 45 have 1 & 3 unrecoverable sectors),
> so, if you can prove me SSD is as good as these, why not…

My question is this: Would a ten year old disk be worth the SATA port
it consumes? Would it be worth its power demands and heating? Ten years
ago, 500GB was a Big Friggin Deal (tm). Today, any fool can go to any
store and get a 4TB drive for less than $200. If you do stuff that
requires disk space, keeping an old drive alive becomes moot after a
certain amount of time. I have all sorts of <100GB drives from ten
years ago: I just take them out to the driveway and do my 20oz hammer
based data wipe.

Here's a related true story. Reading the newspaper in 1994, I saw
Egghead Software was selling a 1000MB drive for $799.00. I figured such
a great price for such amazing technology must be a misprint, but
called them anyway. Yes, they were selling 1000MB for $799.00. I made
them promise to hold one for an hour, RAN to my car, broke every speed
law getting to Egghead, walked in, outwardly casually and inwardly
figuring this can't be true, asked for the drive, handed them my credit
card, and got charged $799.00. Still figuring there'd been a mistake
and I'd be arrested for shoplifting on the way out, I ambled out, got
in my car, carefully drove 2 blocks away, stopped the car, and did a
victory dance.

3 years later I bought an IBM Deskstar drive with 6 times the capacity
for $300.00 :-)

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623165904.40561cf2@mydesk

Re: asciidoc and emacs

[Hendrik Boom]

On Mon, 23 Jun 2014 19:50:31 +, Hendrik Boom wrote:

> The package emacs-goodies-el contains markdown-mode, which is for
> editing markdown files.
> 
> Has anything analogous been packaged foe asciidoc instead?
> 
> If not, there seems to be an asciidoc.el file at
> http://www.emacswiki.org/
> emacs/asciidoc.el   Is there someplace I should put it in my Debian
> testing system where emacs will find it but it won't interfere with any
> of the filesystems that are managed by Debian's package manager?

Actually, asciidoc.el seems to be a set of commands for editing asciidoc, 
not a mode for doing syntax coloring and the like.  Probably not what I 
want.

-- hendrik


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/loa2np$2bv$3...@ger.gmane.org

asciidoc and emacs

[Hendrik Boom]

The package emacs-goodies-el contains markdown-mode, which is for editing 
markdown files.

Has anything analogous been packaged foe asciidoc instead?

If not, there seems to be an asciidoc.el file at http://www.emacswiki.org/
emacs/asciidoc.el   Is there someplace I should put it in my Debian 
testing system where emacs will find it but it won't interfere with any 
of the filesystems that are managed by Debian's package manager?

-- hendrik



-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/loa0e7$2bv$2...@ger.gmane.org

Re: Chromium cannot access pages.

[Hendrik Boom]

On Wed, 18 Jun 2014 08:15:06 -0400, The Wanderer wrote:

> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA512
> 
> On 06/18/2014 04:45 AM, Florian Ernst wrote:
> 
>> On Tue, Jun 17, 2014 at 05:30:44PM +0200, Sven Joachim wrote:
>> 
>>> On 2014-06-17 14:39 +0200, Florian Ernst wrote:
> 
 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=751294
 
 35.0.1916.153-2 contains a fix
>>> 
>>> Yes.
>>> 
 and will soon migrate to testing.
>>> 
>>> No, it won't because it FTBFS on amd64.  Based on previous
>>> experience¹, it's likely to take weeks before a fixed package reaches
>>> testing. :-(
>> 
>> Ah, true, I failed to notice that. Yikes, "ld terminated with signal
>> 9", indicating external trouble rather that a problem with the source
>> per se.
> 
> At first blush, that reminds me of bug 751278, in which ld from the
> currently-packaged binutils crashes fairly reliably under some
> circumstances.
> 
> This seems to be fixed in the binutils now in unstable, so probably the
> problem will go away fairly soon.
> 
> - --
>The Wanderer
> 

However, switching my sources.list to sid temporarily worked just fine.

-- hendrik


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/loa03m$2bv$1...@ger.gmane.org

Re: Progress on my new Debian box

[Bzzzz]

On Tue, 24 Jun 2014 04:34:00 +1000
Andrew McGlashan  wrote:

> I believe that would be true of quite /old/ SSD drives, but
> definitely not for newer ones. 

I wouldn't be so positive… until a real independent lab,
conducting real tests (especially with a high number of
small files, test curiously (much too) often absent from
"testers" sites).

> The new drives are subject to write issues,

Yeah, like older ones.

> but to hit that problem will take just as long as a
> traditional spinning drive -- they too have limits, spinning
> drives are mechanical.

May be, but most of my disks have a ≥ 10 years life (24/7) with
a very few errors (only 2 of 45 have 1 & 3 unrecoverable sectors),
so, if you can prove me SSD is as good as these, why not…

> There have been very heavy torture tests on thew newer range of
> SSDs and they are performing exceptionally well with mega data
> being written [1], up to fairly heavy data usage levels.

All are biased ("strangely", to lower the write errors due to
multiple write repetitions on the same sectors); this is why
until a _real_ lab, with plausible tests protocols and 
methodology doesn't make a test, I won't trust it more than
my first underwear :)
 
> There is apparently a way to restore SSD drives to original
> condition by super heating the layer that breaks down (due to
> writes), targeting the exact spot with the right temperature
> returns the SSD drive to brand new state.  Not sure when this
> newest generation will hit the market though. [2]

Yeah, go figure heating _some_ cells among all in a today's
chip density; not to mention that I don't see other 
sites/labs/researchers saying the same thing.

On this ground too, us firms can't be trusted as they hire
and pay indelicate specialists to _get_ the result they 
_want_; just as monsanto or the govts does.

-- 
ptinou: the only thing that surprised me with vi$ta
ptinou: was when it told me it was going to deactivate my
keyboard to improve the stability of my system


signature.asc
Description: PGP signature

Re: apt-get issues with download speed and server access

[Bob Proulx]

Joel Rees wrote:
> Last week, when the kernel update came down the pipe, most of the
> packages had decent speed, but the kernel and some others were cut
> down to about a fifth normal speed.

What archive name are you using?  I am in the US and use
ftp.us.debian.org and when I do I am actually using one of several
servers.

  $ host -t a ftp.us.debian.org
  ftp.us.debian.org has address 64.50.236.52
  ftp.us.debian.org has address 64.50.233.100
  ftp.us.debian.org has address 128.61.240.89

  $ host -t a ftp.us.debian.org | awk '{print$NF}' | xargs -L1 host -t ptr
  100.233.50.64.in-addr.arpa domain name pointer ftp-nyc.osuosl.org.
  89.240.61.128.in-addr.arpa domain name pointer debian.gtisc.gatech.edu.
  52.236.50.64.in-addr.arpa domain name pointer ftp-chi.osuosl.org.
  
That is for today.  The records are routinely updated as mirrors come
and go.  Sometimes a mirror will have problems and need to be dropped
out.  That will be discussed on the debian-mirrors list.

Normally apt will round-robin among the servers in the list.
Sometimes one server will be having problems and will be slow.
Perhaps it is saturating its network connection.  Perhaps it is
suffering a denial of service attack.  Your experience sounds like
one of the mirrors at that time was likely suffering.

Alternatively there is http.debian.net.  It is a CDN.  It is a way to
use a redirector to dynamically select an appropriate mirror.  I have
been using it and it has been working well for me.  See this for
documentation.

  http://http.debian.net/

> Yesterday or Saturday, when I ran apt-get update, I got server access
> errors on wheezy, including backports and security. Ran apt-get clean
> and the access errors went away.

"apt-get clean" simply purges the downloaded .deb files.  It shouldn't
change what you saw.  But "apt-get update" will have an effect.  I
think one "apt-get update" failed due to the mirror selection but then
the next one succeeded.  Maybe.  I think it likely.

> Today, I get access errors again, and they go away after apt-get
> clean. And all packages are downloading at about a fifth my max speed
> from the provider.

It is also possible that the routes over the Internet through your ISP
are asymmetrical at this time.  You could "traceroute" to each of the
mirrors you are using in turn and see how the routes are different.
Look at the times.  Use ping to check each.  It isn't unusual to find
routers in a path that are sick and troubling.

There is a Debian package netselect that can be used to probe
different archive servers.

  apt-cache show netselect

  http://github.com/apenwarr/netselect

> I'm feeling a little paranoid about this. Should I just assume my
> wan-side connection is getting saturated for some reason?

I wouldn't be worried about security because Debian releases are
cryptographically signed.  But I would probe and try to understand the
network slowness.

Bob


signature.asc
Description: Digital signature

Re: Progress on my new Debian box

[Brian]

On Mon 23 Jun 2014 at 12:43:19 -0400, Steve Litt wrote:

> Anyway, things are going well. My other observation is that Debian is
> less like Ubuntu than it initially looks.

Debian looks less like Ubuntu than you originally thought? What did you
expect and why should it matter? - they are two different OSs. Related
maybe; but peas and beans have a lot in common.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623190507.gi29...@copernicus.demon.co.uk

Re: raid/mdadm help

[Bob Proulx]

Rodney D. Myers wrote:
> Linux-Fan wrote:
> > Rodney D. Myers wrote:
> > > Is there a debian specific mail list or online forum to get some
> > > help for a newbie setting up a raid storage device?
> > > 
> > > I can find plenty of tips to get started, but things are not doing
> > > the exact thing(s) the online guides are showing.

It is all about using Debian so this mailing list is good.  Lots of us
are using raid.

> > Also, if you manually want to create a RAID 1 of two partitions use
> > 
> > # mdadm --create /dev/md0 --level=1 --raid-devices=2 \
> > /dev/sda1 /dev/sdb1
> 
> Not attempting to install on a raid, but create a raid for storage

Linux-Fan's suggestion is a good one.  If you are not using it for the
system then it is easier to get experience building raid yourself as
in the above example.  François's suggestions were good too.

This is where I suggest that you think about using LVM in this
combination too.  If your disks are large, and today's disks can be
very large, then I suggest setting up LVM on them and allocating
smaller portions at a time to the file system.  It is much more
flexible than, say, having a 3T filesystem all in one place.

I would play with the above in order to get experience with raid.
Just setting things up, creating file systems, mounting them, checking
status.  Then unmount and wipe clean and then try a different
configuration.  Repeat a few times before you put it into service so
that you can understand what is happening.

To use LVM in the above start as suggested to create /dev/md0.  Then
set it up for lvm with something like this:

  Create a physical volume from the raid.  (It places a signature there.)
pvcreate /dev/md0

  Create a volume group using the physical volumes just created.
vgcreate vg0 /dev/md0

  The status of the above can be seen using the status commands.
  This will show summaries of space usage.
pvs
vgs

  Allocate 100G to a logical volume to be used for a file system.
lvcreate -L100G -nfirst vg0

  Make a file system on it.  (/dev/vg0/first is same as /dev/mapper/vg0-first)
mkfs -t ext4 /dev/vg0/first

  Mount it.
mkdir /srv/first
mount /dev/vg0/first /srv/first

I am just using "first" as an example.  Assuming there will be a
"second".  Name them according to your needs.  I name mine after the
task.  I have "root", "home", "var", "audio", "bak1", and so forth.

I always keep some disk space in reserve.  If a partition needs more
space then it can be expanded online on the fly.

  Extend a logical volume.
vgs
lvextend -L+25G /dev/vg0/first
resize2fs /dev/vg0/first

As long as you have free space available you can easily expand file
systems.  As we recently discussed in this list a few days ago
shrinking a file system is not as simple.  The simple advice is avoid
needing to shrink by planning ahead.  But expanding one is easy and
reliable.

I advice to play with different configurations while the disks are new
and unused to gain experience.  Easy to play now while there isn't
anything on the disks and you can try different things.  Then choose a
configuration that works for you and move forward with it in real use.

Bob


signature.asc
Description: Digital signature

Re: Progress on my new Debian box

[Sven Hartge]

Steve Litt  wrote:
> On Tue, 24 Jun 2014 04:34:00 +1000
> Andrew McGlashan  wrote:

>> There have been very heavy torture tests on thew newer range of SSDs
>> and they are performing exceptionally well with mega data being
>> written [1], up to fairly heavy data usage levels.

>> [1]
>> http://www.pcper.com/news/Storage/continuing-Tech-Report-SSD-torture-test

> There's another SSD limitation: Deleted sections (I don't know the
> official unit) don't necessarily become available for re-use by the
> OS, unless you trim either with fstrim on a periodic basis, or by
> mounting it with the "discard" attribute (but only if the drive
> supports trim, otherwise, lost data). This is kinda sorta like issues
> with fragmentation on the old MSDOS drive format.

Earlier Linux kernel had known problems with the "discard" mount option,
some earlier drives also had a very inferior implementation of TRIM,
resulting in very poor performance.

But all this has been fixed, both on the side of the Linux kernel (no
longer issueing TRIM for every deleted sector but queueing multiple
TRIMs into one big) and on the firmware side of the SSDs.

I see no problem in using "discard" with a recent kernel and a modern
SSD.

Grüße,
Sven.

-- 
Sigmentation fault. Core dumped.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/5apcif5i1...@mids.svenhartge.de

Re: raid/mdadm help

[Pascal Hambourg]

François Patte a écrit :
> 
> # blkid /dev/md0
> 
> /dev/md0: UUID="41Js9Q-0WaZ-JGYR-r88a-FiqA-XS5F-0XWNOy" (of course the
> uuid will be different for you)
> 
> 2- add this in /etc/fstab:
> 
> UUID=41Js9Q-0WaZ-JGYR-r88a-FiqA-XS5F-0XWNOy /storage-raid ext4 rw 0 0

Unlike disks and their partitions, RAID arrays have persistent device
names so why bother to use UUIDs ?


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a8764c.70...@plouf.fr.eu.org

Re: Progress on my new Debian box

[Andrew McGlashan]

btw It looks like exim4 first entered into Debian with Sarge (Debian
3.1, released 2005), here's a link with more info:

http://www.debian-administration.org/article/98/Upgrading_from_Woody_to_Sarge_Part_4_-_Apache2

Cheers
A.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a8792d.9000...@affinityvision.com.au

Re: Progress on my new Debian box

[Andrew McGlashan]

Steve,

This page was last modified in 2006 and it too talks about "latest Debian"

curl -I http://blog.edseek.com/~jasonb/articles/exim4_courier/exim4.html
HTTP/1.1 200 OK
Date: Mon, 23 Jun 2014 18:54:55 GMT
Server: Apache/2.2.16 (Debian)
Last-Modified: Fri, 19 May 2006 19:30:55 GMT
ETag: "11c128-5aba-4142930fdd9c0"
Accept-Ranges: bytes
Content-Length: 23226
Vary: Accept-Encoding
Content-Type: text/html

Installation of Exim4 from the latest Debian GNU/Linux packages
is easy...

You've got to be wary of information / tutorial pages that don't specify
actual versions of Debian ... if they just say Stable or Old-Stable,
then you need to know when it was written or last updated (for fixes).

Cheers
A.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a8782a.3020...@affinityvision.com.au

Re: Progress on my new Debian box

[Sven Hartge]

Andrew McGlashan  wrote:
> On 24/06/2014 3:16 AM, B wrote:
>> On Mon, 23 Jun 2014 12:43:19 -0400
>> Steve Litt  wrote:
 
>>> I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
>>> 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
>>> and swap partition,
>> 
>> As a SSD has limited write capacities, people usually avoid using  it
>> for things that are often (re)written.  Unfortunately, you just
>> indicate _all_ wrong directories to store on a SSD…

> I believe that would be true of quite /old/ SSD drives, but definitely
> not for newer ones.  The new drives are subject to write issues, but
> to hit that problem will take just as long as a traditional spinning
> drive -- they too have limits, spinning drives are mechanical.

> There have been very heavy torture tests on thew newer range of SSDs
> and they are performing exceptionally well with mega data being
> written [1], up to fairly heavy data usage levels.

Right.

Last paragraphs from that TechReport paper:

,
| Given our limited sample size, I wouldn't read too much into exactly how
| many writes each drive handled. The more important takeaway is that all
| of the SSDs, including the 840 Series, performed flawlessly through
| hundreds of terabytes. A typical consumer won't write anything close to
| that much data over the useful life of a drive.
|
| Even with only six subjects, the fact that we didn't experience any
| failures until after 700TB is a testament to the endurance of modern
| SSDs. So is the fact that three of our subjects have now written over a
| petabyte. That's an astounding total for consumer-grade drives, and the
| Corsair Neutron GTX, Samsung 840 Pro, and compressible Kingston HyperX
| 3K are still going!
`

Grüße,
Sven.

-- 
Sigmentation fault. Core dumped.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/3apchvdi1...@mids.svenhartge.de

Re: Progress on my new Debian box

[Steve Litt]

On Tue, 24 Jun 2014 04:34:00 +1000
Andrew McGlashan  wrote:

> On 24/06/2014 3:16 AM, B wrote:
> > On Mon, 23 Jun 2014 12:43:19 -0400
> > Steve Litt  wrote:
> > 
> >> I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
> >> 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
> >> and swap partition,
> > 
> > As a SSD has limited write capacities, people usually avoid
> > using  it for things that are often (re)written.
> > Unfortunately, you just indicate _all_ wrong directories to 
> > store on a SSD…
> 
> I believe that would be true of quite /old/ SSD drives, but definitely
> not for newer ones.  The new drives are subject to write issues, but
> to hit that problem will take just as long as a traditional spinning
> drive -- they too have limits, spinning drives are mechanical.
> 
> There have been very heavy torture tests on thew newer range of SSDs
> and they are performing exceptionally well with mega data being
> written [1], up to fairly heavy data usage levels.
> 
> There is apparently a way to restore SSD drives to original condition
> by super heating the layer that breaks down (due to writes),
> targeting the exact spot with the right temperature returns the SSD
> drive to brand new state.  Not sure when this newest generation will
> hit the market though. [2]
> 
> [1]
> http://www.pcper.com/news/Storage/continuing-Tech-Report-SSD-torture-test
> 
> [2]
> http://www.tomshardware.com/news/flash-nand-dead-heat-heals,19491.html

Thanks Andrew,

I think the issue you're talking about is drive wear and tear, and like
you say, if your drive lasts 3 years, it's served its purpose and you
can buy double the space for the same price you paid for it three years
ago.

There's another SSD limitation: Deleted sections (I don't know the
official unit) don't necessarily become available for re-use by the OS,
unless you trim either with fstrim on a periodic basis, or by mounting
it with the "discard" attribute (but only if the drive supports trim,
otherwise, lost data). This is kinda sorta like issues with
fragmentation on the old MSDOS drive format.

If there comes a time when you want to return the whole drive to
section by section virginity, you can do that as follows:

https://wiki.archlinux.org/index.php/SSD_Memory_Cell_Clearing

However, this sounds like a dangerous operation for a lot of reasons,
so while this new Debian computer is my daily driver, I'll content
myself with occasional fstrim -v / commands.

Thanks,

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623145120.542685ce@mydesk

Re: Progress on my new Debian box

[Sven Hartge]

B  wrote:> 

> On Mon, 23 Jun 2014 12:43:19 -0400
> Steve Litt  wrote:

>> I've got the hardware all set up. AMD dual core 4100, 16GB RAM, 240GB
>> SSD for /, 750GB Western Digital Black for /var, /tmp, /run, and swap
>> partition,

> As a SSD has limited write capacities, people usually avoid using  it
> for things that are often (re)written.  Unfortunately, you just
> indicate _all_ wrong directories to store on a SSD…

I beg to differ. (I know, you already commented on your error while
reading Steves sentence.)

Modern SSDs are no more fragile than a normal 2,5" spinning drive or
even less, since they have no problems with vibration or sudden G
shocks.

Have a look at
http://techreport.com/review/26523/the-ssd-endurance-experiment-casualties-on-the-way-to-a-petabyte

,
| We started with six SSDs: the Corsair Neutron GTX 240GB, Intel 335
| Series 240GB, Samsung 840 Series 250GB, Samsung 840 Pro 256GB, and two
| Kingston HyperX 3K 240GB.
| [...]
| The last time we checked in, the SSDs had just passed the 600TB mark
| [URL1].  They were all functional, but the 840 Series was burning
| through its TLC cells at a steady pace, and even some of the MLC drives
| were starting to show cracks. We've now written over a petabyte, and
| only half of the SSDs remain. Three drives failed at different
| points—and in different ways—before reaching the 1PB milestone.
`

URL1 
http://techreport.com/review/26058/the-ssd-endurance-experiment-data-retention-after-600tb

Of course, if you heavily abuse your SSD (like the techreport guys are
doing it), you will have a bad time.

But you can just use the drive like a normal hard drive, no need to
specially protect them like a raw egg, aside from leaving a bit of space
(10%) unused so the controller has more room for its wear leveling
algorithm.

My own data points for the Samsung 840 256GB Evo in my laptop look like this.

5505 hours running (~230 days) with ~3450GiB written. Wear_Leveling_Count
is at 98, Reallocated_Sector_Ct is at 100.

,
| ID# ATTRIBUTE_NAME  VALUE WORST THRESH TYPE  RAW_VALUE
|   5 Reallocated_Sector_Ct   100   100   010Pre-fail  0
|   9 Power_On_Hours  098   098   000Old_age   5505
|  12 Power_Cycle_Count   099   099   000Old_age   430
| 177 Wear_Leveling_Count 098   098   000Pre-fail  18
| 179 Used_Rsvd_Blk_Cnt_Tot   100   100   010Pre-fail  0
| 181 Program_Fail_Cnt_Total  100   100   010Old_age   0
| 182 Erase_Fail_Count_Total  100   100   010Old_age   0
| 183 Runtime_Bad_Block   100   100   010Pre-fail  0
| 187 Uncorrectable_Error_Cnt 100   100   000Old_age   0
| 190 Airflow_Temperature_Cel 065   052   000Old_age   35
| 195 ECC_Error_Rate  200   200   000Old_age   0
| 199 CRC_Error_Count 100   100   000Old_age   0
| 235 POR_Recovery_Count  099   099   000Old_age   12
| 241 Total_LBAs_Written  099   099   000Old_age   7233847674
`

There is the Swap space, /var and /tmp on that SSD and I suspend-to-disk
the laptop every morning and evening.

I still expect this drive to outlive the laptop it is currently inside
by far.

You also can quite precisely pinpoint the moment a SSD will fail and
backup your data before while a spinning harddrive may fail without
warning at any moment.

I would advice Steve to put Swap and /var back onto the SSD as those are
the areas which gain the most from being on a fast storage medium. 

/tmp can be a tmpfs, which will be written to the Swap if needed, also
profiting from the fast SSD.

/run should already be a tmpfs in Debian Wheezy.

/home can then be on the WD because of the big space available there.

Grüße,
Sven.

-- 
Sigmentation fault. Core dumped.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/2apcgfti1...@mids.svenhartge.de

Re: Progress on my new Debian box

[Andrew McGlashan]

What are you talking about Steve?

On 24/06/2014 2:43 AM, Steve Litt wrote:
> The biggest problem I'm having right now is email. I set up a
> preliminary Dovecot on the Debian machine. For some reason, the only
> way to get the dovecot executable on Wheezy is to follow these
> instructions on adding to your sources list:
> 
> http://wiki2.dovecot.org/HowTo/DebianStable

Umm, that reference is *really* old.


On Debian Stable, woody with exim 3



Woody was Debian 3.0, with end of support 2006 and I don't even remember
using exim before exim4.


This is from Debian current 7.5 Stable:

# apt-cache search dovecot
ciderwebmail - IMAP webmail service
dovecot-common - Transitional package for dovecot
dovecot-core - secure mail server that supports mbox, maildir, dbox and
mdbox mailboxes
dovecot-dbg - debug symbols for Dovecot
dovecot-dev - header files for the dovecot mail server
dovecot-gssapi - GSSAPI authentication support for Dovecot
dovecot-imapd - secure IMAP server that supports mbox, maildir, dbox and
mdbox mailboxes
dovecot-ldap - LDAP support for Dovecot
dovecot-lmtpd - secure LMTP server for Dovecot
dovecot-managesieved - secure ManageSieve server for Dovecot
dovecot-mysql - MySQL support for Dovecot
dovecot-pgsql - PostgreSQL support for Dovecot
dovecot-pop3d - secure POP3 server that supports mbox, maildir, dbox and
mdbox mailboxes
dovecot-sieve - sieve filters support for Dovecot
dovecot-solr - Solr full text search support for Dovecot
dovecot-sqlite - SQLite support for Dovecot
dovecot-antispam - Dovecot plugins for training spam filters
maildrop - mail delivery agent with filtering abilities
mysqmail-dovecot-logger - real-time logging system in MySQL - Dovecot
traffic-logger
phamm - PHP front-end to manage virtual services on LDAP - main package
postfixadmin - Virtual mail hosting interface for Postfix
pysieved - managesieve server
roundcube-plugins-extra - skinnable AJAX based webmail solution - extra
plugins
vmm - manage mail domains/accounts/aliases for Dovecot and Postfix


And this from old-stable:

# apt-cache search dovecot
dovecot-common - secure mail server that supports mbox and maildir mailboxes
dovecot-dbg - debug symbols for Dovecot
dovecot-dev - header files for the dovecot mail server
dovecot-imapd - secure IMAP server that supports mbox and maildir mailboxes
dovecot-pop3d - secure POP3 server that supports mbox and maildir mailboxes
mysqmail-dovecot-logger - real-time logging system in MySQL - Dovecot
traffic-logger
phamm - PHP front-end to manage virtual services on LDAP - main package
pysieved - managesieve server


I've got the following installed on my mail server (old-stable), there
were no special tricks -- except for some local configuration
adjustments I wanted to make:

# dpkg-query -l|grep dovec
ii  dovecot-common   1:1.2.15-7   secure
mail server that supports mbox and maildir mailboxes
ii  dovecot-imapd1:1.2.15-7   secure
IMAP server that supports mbox and maildir mailboxes
ii  dovecot-pop3d1:1.2.15-7   secure
POP3 server that supports mbox and maildir mailboxes


Cheers

-- 
Kind Regards
AndrewM

Andrew McGlashan
IT Support & Broadband Solutions

Current Land Line No: 03 9012 2102
Mobile: 04 2574 1827 Fax: 03 9012 2178

Affinity Vision Australia Pty Ltd
http://affinityvision.com.au
https://securemywireless.com.au
https://adsl2choice.net.au

In Case of Emergency --  http://affinityvision.com.au/ice.html


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a8761b.5080...@affinityvision.com.au

Re: Getting rights right

[Bob Proulx]

Diogene Laerce wrote:
> Hi Bob,

Hi!

> First thank you for the detailed answer, you kind of preventively
> answered to all my doubts or interrogations. :)

Yay!  Then I was successful!  :-)  \o/

> I try to set up a new line of security (files and network) as I just
> changed country and instead of being in one mostly targeting others, I
> am now in one mostly targeted by others. :D

fun!

> >   chmod -R u+rwX,go-rwx /home/user/Documents
> 
> I ran this command to restart the process :
>  find /home/user/Documents -type f -exec chmod u+rw,go-rwx -R {} \;
> and will make executable all following files according the needs.

More comments from me about the above.  It is pretty good.  It doesn't
do anything bad.  But it could be better.

  find $directory -type f

That will find all files below the specified directory.

  -exec chmod u+rw,go-rwx -R {} \;

That will chmod each file (each due to "{} \;") to the specified
symbolic mode.  All good.

The -R is a little odd there.  That says to recursively change files
down a directory hierarchy.  Of course the find is only going to pass
it files so there won't ever be a directory seen.  The -R in that case
isn't doing any harm but neither is it doing anything at all.  Also
'find' is already the super powerful nice recursive command.  It is
the biggest and best tool in the toolbox.  Since recursive commands
can get away from people sometimes I think it best to use one of them
at a time.  :-)

The "{} \;" part is the traditional way to do -exec and you will find
it in many Unix text books forever.  It has some disadvantages though.
It invokes the command once for each file.  That isn't as efficient as
it could be.  More than a decade ago find was enhanced to include the
"{} +" construct as a new and better form of this.  For one "+" isn't
special to the shell and does not need to be escaped.  That is good by
itself.  But "{} +" also invokes the command once and passes the
entire argument list, or as much of the argument list as possible on
the system (it is system dependent), to the command.  Therefore it is
much more efficient since it reduces the number of fork and exec calls
and makes handling the large file lists more efficient.

If we polish up your command just a tiny bit we have this:

  find /home/user/Documents -type f -exec chmod u+rw,go-rwx {} +

Again, your original command is fine and got the job done.  I just
wanted to polish it up a small amount for next time.

> > That is usually called UPG (User Private Group).
> >...
> 
> After reading this, I actually found that :
> 
> umask and level of security : The umask command be used for setting
> different security levels as follows:
> 
> umask value   Security level  Effective permission (directory)
> 022   Permissive  755
> 026   Moderate751
> 027   Moderate750
> 077   Severe  700
> 
> in there :
> http://www.cyberciti.biz/tips/understanding-linux-unix-umask-value-usage.html

I gave that a quick skim and that article seems factually accurate.
However trying to assign human fuzzy names "Permissive, Moderate,
Severe" to them is completely arbitrary and I disagree with the
direction at that point.  I would rather have features and
capabilities line up with the particular goals to be accomplished.

Because frankly I would say for "Severe" security that I would turn
the power off!  That would be severe! :-)  Otherwise it is just
different settings for different features.

> And I was planning to set a "severe" security plan. Based on the
> thinking that I have 3 computers (that only I use) to run behind a box
> and that I thought wiser to set them to the maximum security first, find
> out what they will exchange in second and then update the permissions
> accordingly, as I have very little impact on the box security.

Given all of the above I think that is a reasonable plan.  I can't
argue with the direction of your thinking.  But I also understand how
these permissions work and how they interact.  So I personally
wouldn't be recommending "Severe".  I recommend a UPG "umask 02" which
isn't even an option from the above list.  If you are a sole user on
your own system then it doesn't really matter.

> I then opted for the umask 077. I'm not sure if it's really justified
> but it couldn't do no harm.. I guess. :)

But for example if you share files by making tar files and sending
them out then that "Severe" setting creates problems for others when
they unpack the files and the settings are propagated to them.  I
wouldn't make a software release bundle that way for example.  Also
for example if you interacted with others through a version control
server then permissions can leak through there too.  Again it all
depends upon what you are doing and how you are interacting with
others.  I am not saying you are doing any of those things but I think
eventually you will want to share some files with someone and then you
need to be aware of the file permis

Re: Progress on my new Debian box

[Steve Litt]

On Mon, 23 Jun 2014 18:33:27 + (UTC)
Curt  wrote:

> On 2014-06-23, Tom Furie  wrote:
> >
> > Try again. The "often (re)written" directories are on a WD spinning
> > disk. His data (not sure if that includes homes, but if not I would
> > strongly advise it) is on yet another spinning disk.
> >
> 
> Right, but as his data appears to be on a WD "Green" maybe Steve
> should watch out for the infamous Load Cycle Count bug, if applicable.

Thanks Curt,

The WD green I bought a couple months ago didn't have the Load Cycle
Count Bug, so I'm temporarily assuming this one doesn't either. If it
does turn out to have that, I'll just replace it with a (hotter) black.

My current case is *much* cooler and more air-conductive and has much
more fannage than my old one, so one or two WD blacks can probably run
just fine without those lame "hard disk coolers".

Thanks,

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623144109.72f7b4a5@mydesk

Re: Progress on my new Debian box

[Curt]

On 2014-06-23, Tom Furie  wrote:
>
> Try again. The "often (re)written" directories are on a WD spinning
> disk. His data (not sure if that includes homes, but if not I would
> strongly advise it) is on yet another spinning disk.
>

Right, but as his data appears to be on a WD "Green" maybe Steve should
watch out for the infamous Load Cycle Count bug, if applicable.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/slrnlqgskq.22f.cu...@einstein.electron.org

Re: Progress on my new Debian box

[Andrew McGlashan]

On 24/06/2014 3:16 AM, B wrote:
> On Mon, 23 Jun 2014 12:43:19 -0400
> Steve Litt  wrote:
> 
>> I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
>> 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
>> and swap partition,
> 
> As a SSD has limited write capacities, people usually avoid
> using  it for things that are often (re)written.
> Unfortunately, you just indicate _all_ wrong directories to 
> store on a SSD…

I believe that would be true of quite /old/ SSD drives, but definitely
not for newer ones.  The new drives are subject to write issues, but to
hit that problem will take just as long as a traditional spinning drive
-- they too have limits, spinning drives are mechanical.

There have been very heavy torture tests on thew newer range of SSDs and
they are performing exceptionally well with mega data being written [1],
up to fairly heavy data usage levels.

There is apparently a way to restore SSD drives to original condition by
super heating the layer that breaks down (due to writes), targeting the
exact spot with the right temperature returns the SSD drive to brand new
state.  Not sure when this newest generation will hit the market though. [2]

[1]
http://www.pcper.com/news/Storage/continuing-Tech-Report-SSD-torture-test

[2] http://www.tomshardware.com/news/flash-nand-dead-heat-heals,19491.html


Cheers
A.



signature.asc
Description: OpenPGP digital signature

Re: chkrootkit message

[Brian]

On Mon 23 Jun 2014 at 19:56:15 +0400, Reco wrote:

> On Mon, Jun 23, 2014 at 10:03:30AM +0200, François Patte wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA1
> > 
> > Bonjour,
> > 
> > I get this alert message (concerning lightdm) from chkrootkit
> > 
> > ! RUID  PID TTYCMD
> > ! root 3153 tty7   /usr/bin/X :0 -seat seat0 -auth
> > /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch
> > 
> > What does it mean?
> 
> A false positive. See this, for example:
> 
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=677315

There is no well-documented case of chrootkit ever giving a true
positive; false positives are its stock in trade. What do you expect of
a program which searches for things which do not exist or which have no
relevance (if they ever had) on a modern Linux?

Clapping loudly is very effective at keeping elephants out of my garden. :)
What use is chkrootkit?

(Yes, I know it doesn't answer the question, but my response could lead
to a mass purging of chkrootkit from users' systems :) ).


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623182815.gh29...@copernicus.demon.co.uk

Re: apt-get issues with download speed and server access

[Brian]

On Mon 23 Jun 2014 at 12:04:55 -0400, Steve Litt wrote:

> I don't know if any conclusions can be drawn from my anecdotes, but if
> you collect enough anecdotes perhaps it can help you figure out what's
> going wrong.

Lots of conclusions can be drawn from anecdotal evidence but a proven
aspect of troubleshooting is to rely only on factual evidence. The
conclusions drawn from factual evidence may be be incorrect but at least
there is a verifiable, repeatable framework to return to. Anecdotal
evidence has a habit of morphing over time, which, amongst other things,
makes it useless for figuring out anything.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623182733.gg29...@copernicus.demon.co.uk

Re: Progress on my new Debian box

[Steve Litt]

On Mon, 23 Jun 2014 18:26:48 +0100
Tom Furie  wrote:

> On Mon, Jun 23, 2014 at 07:16:27PM +0200, B wrote:
> > On Mon, 23 Jun 2014 12:43:19 -0400
> > Steve Litt  wrote:
> > 
> > > I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
> > > 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
> > > and swap partition,
> > 
> > As a SSD has limited write capacities, people usually avoid
> > using  it for things that are often (re)written.
> > Unfortunately, you just indicate _all_ wrong directories to 
> > store on a SSD…
> 
> Try again. The "often (re)written" directories are on a WD spinning
> disk. His data (not sure if that includes homes, but if not I would
> strongly advise it) is on yet another spinning disk.
> 
> Cheers,
> Tom

Yes, I should clarify. The following is the contents of my mount
command:

http://paste.debian.net/106401

The one with a UUID device, mounted to /, is really /dev/sda, the
SSD. /dev/sdb is the 3TB WD green for my data, and /dev/sdc is the 750GB
WD black, which is faster than the green, is for the changing part of
the OS and for my swap partition.

The idea was that I rely on my own discipline not to put any frequently
changing stuff on /dev/sda.

By the way, the following command shows that my SSD supports trim:

hdparm -I | grep -i trim

And so I can use the following command every once in a while to put no
longer used disk sections back in the pool:

fstrim -v /

Like mentioned in a previous message, on this box I went cheap rather
than going full-featured, so I anticipate using it only 2 years. I hope
the next one has 64GB RAM, and I hope by that time SSD storage is a lot
cheaper and more forgiving of a lot of writes.

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623141901.2a29c14e@mydesk

Re: Progress on my new Debian box

[Bzzzz]

On Mon, 23 Jun 2014 18:26:48 +0100
Tom Furie  wrote:

> Try again. The "often (re)written" directories are on a WD spinning
> disk. His data (not sure if that includes homes, but if not I would
> strongly advise it) is on yet another spinning disk.

Oops, I missed the comma :(

Manufacturers (or researchers) don't work very much on this
specific point, which is too bad; imagine a new generation
PCIe SSD: ~2GB/s write, ~3GB/s read and an almost infinite
lifetime.

My guess is that it already exists in labs, but will never
be commercialized for programmed obsolescence matters :((

-- 
 Imagine a jewish Santa.
 "Hi kids, what do I sell you?"


signature.asc
Description: PGP signature

Re: Progress on my new Debian box

[Tom Furie]

On Mon, Jun 23, 2014 at 07:16:27PM +0200, B wrote:
> On Mon, 23 Jun 2014 12:43:19 -0400
> Steve Litt  wrote:
> 
> > I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
> > 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
> > and swap partition,
> 
> As a SSD has limited write capacities, people usually avoid
> using  it for things that are often (re)written.
> Unfortunately, you just indicate _all_ wrong directories to 
> store on a SSD…

Try again. The "often (re)written" directories are on a WD spinning
disk. His data (not sure if that includes homes, but if not I would
strongly advise it) is on yet another spinning disk.

Cheers,
Tom

-- 
We fight only when there is no other choice.  We prefer the ways of
peaceful contact.
-- Kirk, "Spectre of the Gun", stardate 4385.3


signature.asc
Description: Digital signature

Re: Progress on my new Debian box

[Bzzzz]

On Mon, 23 Jun 2014 12:43:19 -0400
Steve Litt  wrote:

> I've got the hardware all set up. AMD dual core 4100, 16GB RAM,
> 240GB SSD for /, 750GB Western Digital Black for /var, /tmp, /run,
> and swap partition,

As a SSD has limited write capacities, people usually avoid
using  it for things that are often (re)written.
Unfortunately, you just indicate _all_ wrong directories to 
store on a SSD…

-- 
Wink : 2s, I'll be back in 10 minutes
Mysterius : capitaine Kirk, a temporal singularity starboard!


signature.asc
Description: PGP signature

Re: Progress on my new Debian box

[Lisi Reisz]

On Monday 23 June 2014 17:43:19 Steve Litt wrote:
> My other observation is that Debian is
> less like Ubuntu than it initially looks.

:-))  Quite!!  But I don't actually even think that it initially looks like 
it

Lisi


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/201406231756.43843.lisi.re...@gmail.com

Progress on my new Debian box

[Steve Litt]

Hi all,

I've got the hardware all set up. AMD dual core 4100, 16GB RAM, 240GB
SSD for /, 750GB Western Digital Black for /var, /tmp, /run, and swap
partition, and a 3GB Western Digital Green for all my data.

I installed Debian Wheezy 7.5 network install with "nonfree yes", and
xxxterm (soon to be called xombrero) plays Youtube videos perfectly.
Xfce right now, dwm later. I have UMENU and VimOutliner working
perfectly. Gnumeric and LyX are installed, although I haven't tested
them extensively.

The biggest problem I'm having right now is email. I set up a
preliminary Dovecot on the Debian machine. For some reason, the only
way to get the dovecot executable on Wheezy is to follow these
instructions on adding to your sources list:

http://wiki2.dovecot.org/HowTo/DebianStable

Then, so far, I can't access my new dovecot from my Debian machine,
although I can access it over the wire from my Ubuntu machine.
Thickening the plot even more is that nmap 192.168.100.4 from
192.168.100.4 won't read any open ports, even though the same command
issued from 192.168.100.2 shows 22, 80, 111, 143 and 993 open. I
suspect some sort of firewall issue, although at present I don't know
enough to prove or disprove that.

Anyway, things are going well. My other observation is that Debian is
less like Ubuntu than it initially looks.

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623124319.0fe5fbab@mydesk

Re: Samba4 is missing

[Darac Marjal]

On Mon, Jun 23, 2014 at 06:13:08PM +0200, Jordi Clariana wrote:
>Hello,
>Today I realized that package samba4 is missing from repositories. Is this
>normal?
>I had installed samba4 packages from one or two month ago, and now it is
>unavailable.
># apt-cache search samba4
>libsamba-hostconfig-dev - Samba host configuration library - development
>files
>libsamba-hostconfig0 - Samba host configuration library
>samba4-clients - client utilities from Samba 4
>samba4-common-bin - Samba 4 common files used by both the server and the
>client
>samba4-dev - tools for extending Samba
>samba4-testsuite - test suite from Samba 4
>samba4 package is missing in the search results.
>Thanks.

According to Bug #726642, this is because samba itself has now reached
version 4.



signature.asc
Description: Digital signature

Samba4 is missing

[Jordi Clariana]

Hello,

Today I realized that package samba4 is missing from repositories. Is this
normal?
I had installed samba4 packages from one or two month ago, and now it is
unavailable.

# apt-cache search samba4
libsamba-hostconfig-dev - Samba host configuration library - development
files
libsamba-hostconfig0 - Samba host configuration library
samba4-clients - client utilities from Samba 4
samba4-common-bin - Samba 4 common files used by both the server and the
client
samba4-dev - tools for extending Samba
samba4-testsuite - test suite from Samba 4

samba4 package is missing in the search results.

Thanks.


*Jordi Clariana*
*IT Manager**Senior System Administrator*

ATRAPALO.COM 
Aribau 185, 1º
08021 Barcelona
Tel. directo: 935208446
Tel. oficina: 933193001 ext. 203
Fax. 935208400

Re: raid/mdadm help

[François Patte]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Le 23/06/2014 17:22, Rodney D. Myers a écrit :

which should be still similar in Debian 7. If you are new to RAID,


So what HTH you says is the good way: create one partitions on each of
your storages devices (say sdX1 and sdY1) then use the given command:

# mdadm --create /dev/md0 --level=1 --raid-devices=2 \
/dev/sdX1 /dev/sdY1

Then format the created raid array

# mkfs.ext4 /dev/md0

Create a mount point for your raid:

# mkdir /storage-raid

Try the result

# mount /dev/md0 /storage-raid

If everything ok, finish your install

1-

# blkid /dev/md0

/dev/md0: UUID="41Js9Q-0WaZ-JGYR-r88a-FiqA-XS5F-0XWNOy" (of course the
uuid will be different for you)

2- add this in /etc/fstab:

UUID=41Js9Q-0WaZ-JGYR-r88a-FiqA-XS5F-0XWNOy /storage-raid ext4 rw 0 0

3-

# mdadm --detail --scan >> /etc/mdadm/mdadm.conf


reboot and see if everything works fine...

- -- 
François Patte
UFR de mathématiques et informatique
Laboratoire CNRS MAP5, UMR 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)1 8394 5849
http://www.math-info.univ-paris5.fr/~patte
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlOoVCoACgkQdE6C2dhV2JXV6gCcC9SEIzSgCqvOwTHJuRr466ty
WxIAnjflYxOJDGc9ppw0k70D954QOrEN
=duvV
-END PGP SIGNATURE-


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a8542a.9050...@mi.parisdescartes.fr

Re: Does LXDE really require lightdm?

[Steve Litt]

On Mon, 23 Jun 2014 23:52:37 +0900
Osamu Aoki  wrote:

> Hi,
> 
> On Sat, Jun 21, 2014 at 08:31:50PM -0400, Steve Litt wrote:
> > Hi all,
> > 
> > I installed LXDE on a no-X, no-desktop virgin network Wheezy 64bit
> > install with non-free software allowed, and on the next boot it went
> > into lightdm. The only thing I could find that installed and
> > required lightdm was LXDE. I uninstalled LXDE, installed Xfce,
> > installed whatever bestows startx, and bang, X from the CLI command
> > line, no *dm needed.
> 
> I think you should learn to use aptitude to look-into Debian's
> resources.  

Cool. I'll do that next weekend. I've used apt-get or synaptic until
now, but obviously I need finer granularity.

> Here are the answer by running aptitude.
> 
> > 1) Am I correct that Debian's LXDE package installs lightdm?
> 
> It depends on what yopu mean by "LXDE package".  
> 
> If you mean "task-lxde-desktop", yes it is "depends".
> 
> If you mean "lxde", practically yes since it is "recommends".
> 
> > 2) Does that come from the LXDE project, or is it a Debian thing?
> 
>   Homepage: https://launchpad.net/lightdm
>   http://www.freedesktop.org/wiki/Software/LightDM/

ROFLMAO, I need to improve my English (which is my native language).
What I *meant* to say was "does the *dependency* come from lxde, or
does the *dependency* come from Debian.

> 
> > 3) Is there a way to turn off LXDE's install of lightdm?
> 
> If you chose "lxde", you install without recommends.  That is easy
> with aptitude and apt-gey can do that via command line.  Read the
> manual pages of them.

I'll be doing that next weekend.

> 
> > The whole reason I'm switching from Xubuntu to Debian is to get away
> > from both Plymouth and *dm. Fortunately, I find LXDE desireable,
> > but no way do I find it necessary.
> 
> You can go less with bare Openbox window manager :-)

If you like Openbox, you'll love my ultimate destination: dwm! But when
I'm first installing a computer and getting all the functionalities
working, including hundreds of home-grown shellscripts, python, perl,
ruby and lua programs, I like a user interface that gives me more
context.

Later, when my interface is merely a way to run programs, I switch to
something like Openbox or dwm.

Thanks Osamu,

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623121503.5c23d100@mydesk

Re: apt-get issues with download speed and server access

[Steve Litt]

On Mon, 23 Jun 2014 23:21:48 +0900
Joel Rees  wrote:

> (Seems like I remember seeing a thread on this recently, but I don't
> see it in the last week's posts.)
> 
> Last week, when the kernel update came down the pipe, most of the
> packages had decent speed, but the kernel and some others were cut
> down to about a fifth normal speed.
> 
> Yesterday or Saturday, when I ran apt-get update, I got server access
> errors on wheezy, including backports and security. Ran apt-get clean
> and the access errors went away.
> 
> Today, I get access errors again, and they go away after apt-get
> clean. And all packages are downloading at about a fifth my max speed
> from the provider.
> 
> I'm feeling a little paranoid about this. Should I just assume my
> wan-side connection is getting saturated for some reason?

Hi Joel,

I don't know whether this relates to what you're saying, but a few days
ago debian.org was down, and xubuntu.org was *incredibly* slow, both on
my side and at http://www.isup.me, which told me both were flat out
down. I just went to sleep, and upon waking the next day, both were
doing well.

Also, a couple days ago, using the Debian Wheezy 64bit network install,
the debian.org default mirror didn't work, so I switched to the rit.edu
mirror and it worked perfectly.

I don't know if any conclusions can be drawn from my anecdotes, but if
you collect enough anecdotes perhaps it can help you figure out what's
going wrong.

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623120455.09ece9af@mydesk

Re: chkrootkit message

[Reco]

Hi.

On Mon, Jun 23, 2014 at 10:03:30AM +0200, François Patte wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Bonjour,
> 
> I get this alert message (concerning lightdm) from chkrootkit
> 
> ! RUID  PID TTYCMD
> ! root 3153 tty7   /usr/bin/X :0 -seat seat0 -auth
> /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch
> 
> What does it mean?

A false positive. See this, for example:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=677315

Reco


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623155614.GA14580@x101h

[SOLVED] Re: Getting rights right

[Diogene Laerce]

Thanks to stop by. :)

>> I have a strange behavior lately on my Deby. After a run of :
>>
>> chown user:user -R /home/user/Documents
>>
>> and :
>>
>> chmod 700 -R /home/user/Documents

[...]

> Which commands did you run as root, or su/sudo root?
> 
> I have sometimes, especially when moving files between systems where
> the numeric uids are different, found myself doing a recursive chmod
> or chown, and being unable to change permissions on some files because
> of the user I was running as. Then, after the directory permissions
> changed, being able to change them on a second pass.

I ran all commands with sudo.

Actually I remember to encounter the case also but what did the trick
here was to pass the command in symbolic mode not in octal. I don't why
but all things work just fine now.

I don't know if it is the reason but the recalcitrant files were mostly
(maybe all) windows files. Just a hypothesis..

Cheers
-- 
“One original thought is worth a thousand mindless quotings.”
“Le vrai n'est pas plus sûr que le probable.”

  Diogene Laerce



signature.asc
Description: OpenPGP digital signature

Re: ntp and multiple OSes

[Chris Davies]

Bob Proulx  wrote:
> Ah...  I had not ever seen ntpdate or rdate used for clock comparison
> before.

It really is a very useful tool for clock comparisons.
Chris


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/nfsl7bxjcp@news.roaima.co.uk

Re: raid/mdadm help

[Rodney D. Myers]

On Mon, 23 Jun 2014 16:08:15 +0200
Linux-Fan  wrote:

> On 06/23/2014 03:40 PM, Rodney D. Myers wrote:
> > Is there a debian specific mail list or online forum to get some
> > help for a newbie setting up a raid storage device?
> > 
> > I can find plenty of tips to get started, but things are not doing
> > the exact thing(s) the online guides are showing.
> > 
> > thank you  
> 
> The Debian installer can do it. Compare
> http://www.texsoft.it/index.php?c=hardware&m=hw.storage.boot-raid-squeeze&l=it
> which should be still similar in Debian 7. If you are new to RAID,
> also simulate the case of disk failure and replacement.
> 
> The step ``Install GRUB to all disks'' is better described as
> 
>   # dpkg-reconfigure grub-pc
> 
> Also, if you manually want to create a RAID 1 of two partitions use
> 
>   # mdadm --create /dev/md0 --level=1 --raid-devices=2 \
>   /dev/sda1 /dev/sdb1
> 
> HTH
> Linux-Fan

Not attempting to install on a raid, but create a raid for storage

-- 
Rodney D. Myers 

They that can give up essential liberty to obtain a
little temporary safety deserve neither liberty nor safety.
        Ben Franklin - 1759


signature.asc
Description: PGP signature

Re: Does LXDE really require lightdm?

[Osamu Aoki]

Hi,

On Sat, Jun 21, 2014 at 08:31:50PM -0400, Steve Litt wrote:
> Hi all,
> 
> I installed LXDE on a no-X, no-desktop virgin network Wheezy 64bit
> install with non-free software allowed, and on the next boot it went
> into lightdm. The only thing I could find that installed and required
> lightdm was LXDE. I uninstalled LXDE, installed Xfce, installed
> whatever bestows startx, and bang, X from the CLI command line, no *dm
> needed.

I think you should learn to use aptitude to look-into Debian's
resources.  Here are the answer by running aptitude.

> 1) Am I correct that Debian's LXDE package installs lightdm?

It depends on what yopu mean by "LXDE package".  

If you mean "task-lxde-desktop", yes it is "depends".

If you mean "lxde", practically yes since it is "recommends".

> 2) Does that come from the LXDE project, or is it a Debian thing?

  Homepage: https://launchpad.net/lightdm
  http://www.freedesktop.org/wiki/Software/LightDM/

> 3) Is there a way to turn off LXDE's install of lightdm?

If you chose "lxde", you install without recommends.  That is easy with
aptitude and apt-gey can do that via command line.  Read the manual
pages of them.

> The whole reason I'm switching from Xubuntu to Debian is to get away
> from both Plymouth and *dm. Fortunately, I find LXDE desireable, but no
> way do I find it necessary.

You can go less with bare Openbox window manager :-)

Osamu


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140623145237.GA14440@goofy

Re: Early access to a console (during runlevel 1)

[Bzzzz]

On Sun, 22 Jun 2014 23:52:46 -0600
Bob Proulx  wrote:

>   $ man fsck.ext4

Ok, my bad 'cos I didn't re-read this for a long time,
time where -a was different from -p.

So, as fixes are those that won't need user's touch,
I agree to your argument :)

-- 
BOFH excuse #345:
Having to manually track the satellite.


signature.asc
Description: PGP signature

Re: apt-get update: unnecessary use of disk space

[Bzzzz]

On Sun, 22 Jun 2014 23:24:32 -0600
Bob Proulx  wrote:

> What do you mean when you say "these blocks won't be free ..."
> without defragmenting?  Please explain.  If you have references to
> share that explained the details that would be great.

Just think about this:
* HD original = 1000
* HD -5%. =  950

The -5% can _only_ be used by processes (eg: to write their
log files); so for users, until a part or the whole -5% is
released by re-tuning the disk, the HD will always looks 950.

Now, if you know an underlying mechanism that silently retrocede
place from these 5% to users, please tell me what it is.

Hence "these blocks won't be free", as users only see a HD of 950.

And BTW, even root sees a HD of 950, not 1000 (but only root can
use the remaining 5% it doesn't see).
 
> Also as I understand it use of e2defrag is not recommended.  Using
[cut blurb]

This isn't the question, try not to be silly; I took the example
of defrag pgm to illustrate the fact that the reserved 5% can
only be accessed by such a program, enforcing the fact that these
5% are _unreachable_ from anyone (except root in particular cases).

As either you wanna argue for looong sterile threads or you didn't
understood, I put it in short terms:
* HD (formatted) is 1000 (without any reservation)
* Regular 5% reservation drops its capacity to 950
* Everybody only sees and uses 950, not 1000
* The 5% resa will _never_ be available to users

Conclusion: users see and use a 950 HD, meaning if I/O
slows down when HD is 95% full, it will slow down at:
950 (HD seen capacity) * 95% = 902.5

-- 
 these stupid head hunters want resumes in ms word format
 can you write shit in tex and convert it to word?
 \converttoword{shit}


signature.asc
Description: PGP signature

Re: Getting rights right

[Joel Rees]

Just a shot in the dark --

On Sun, Jun 22, 2014 at 11:58 PM, Diogene Laerce  wrote:
> Hi,
>
> I have a strange behavior lately on my Deby. After a run of :
>
> chown user:user -R /home/user/Documents
>
> and :
>
> chmod 700 -R /home/user/Documents
>
> I run :
>
> find /home/user/Documents ! -perm 0700
>
> But I still get a list of files like :
>
>.
>.
>.
> /home/user/Documents/administrative/passport (2).png
> /home/user/Documents/administrative/00IMG_0006.jpg
> /home/user/Documents/administrative/IMG_0016.jpg
> /home/user/Documents/administrative/visit.appart
>.
>.
>.
>
> Then if I checked their rights with :
>
> ls -la /home/user/Documents/administrative
>
> They are anyway all well checked :
>
> -rw---
>
> Could someone have an idea of what is going on ? What should I believe ?
> The "find" command or the "ls" one ?
>
> Thank you
> --
> “One original thought is worth a thousand mindless quotings.”
> “Le vrai n'est pas plus sûr que le probable.”
>
>   Diogene Laerce
>

Which commands did you run as root, or su/sudo root?

I have sometimes, especially when moving files between systems where
the numeric uids are different, found myself doing a recursive chmod
or chown, and being unable to change permissions on some files because
of the user I was running as. Then, after the directory permissions
changed, being able to change them on a second pass.

-- 
Joel Rees

Be careful where you see conspiracy.
Look first in your own heart.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/caar43ins2ohaarkpijkvf6+5bgtpr-9db1tyzs0x9d25zha...@mail.gmail.com

apt-get issues with download speed and server access

[Joel Rees]

(Seems like I remember seeing a thread on this recently, but I don't
see it in the last week's posts.)

Last week, when the kernel update came down the pipe, most of the
packages had decent speed, but the kernel and some others were cut
down to about a fifth normal speed.

Yesterday or Saturday, when I ran apt-get update, I got server access
errors on wheezy, including backports and security. Ran apt-get clean
and the access errors went away.

Today, I get access errors again, and they go away after apt-get
clean. And all packages are downloading at about a fifth my max speed
from the provider.

I'm feeling a little paranoid about this. Should I just assume my
wan-side connection is getting saturated for some reason?

-- 
Joel Rees

Be careful where you see conspiracy.
Look first in your own heart.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/caar43ipzv9d-svt-pjut62l+rb_o2bxaczd6xgimgaszjmr...@mail.gmail.com

Re: raid/mdadm help

[Linux-Fan]

On 06/23/2014 03:40 PM, Rodney D. Myers wrote:
> Is there a debian specific mail list or online forum to get some help
> for a newbie setting up a raid storage device?
> 
> I can find plenty of tips to get started, but things are not doing the
> exact thing(s) the online guides are showing.
> 
> thank you

The Debian installer can do it. Compare
http://www.texsoft.it/index.php?c=hardware&m=hw.storage.boot-raid-squeeze&l=it
which should be still similar in Debian 7. If you are new to RAID, also
simulate the case of disk failure and replacement.

The step ``Install GRUB to all disks'' is better described as

# dpkg-reconfigure grub-pc

Also, if you manually want to create a RAID 1 of two partitions use

# mdadm --create /dev/md0 --level=1 --raid-devices=2 \
/dev/sda1 /dev/sdb1

HTH
Linux-Fan



signature.asc
Description: OpenPGP digital signature

raid/mdadm help

[Rodney D. Myers]

Is there a debian specific mail list or online forum to get some help
for a newbie setting up a raid storage device?

I can find plenty of tips to get started, but things are not doing the
exact thing(s) the online guides are showing.

thank you

-- 
Rodney D. Myers 

They that can give up essential liberty to obtain a
little temporary safety deserve neither liberty nor safety.
        Ben Franklin - 1759


signature.asc
Description: PGP signature

Re: Getting rights right

[Diogene Laerce]

Hi Bob,

First thank you for the detailed answer, you kind of preventively
answered to all my doubts or interrogations. :)

I try to set up a new line of security (files and network) as I just
changed country and instead of being in one mostly targeting others, I
am now in one mostly targeted by others. :D

>> I have a strange behavior lately on my Deby. After a run of :
>> chown user:user -R /home/user/Documents
>> and :
>> chmod 700 -R /home/user/Documents
> 
> Unfortunately that command was a mistake.  That will set rwx for owner
> on all files unconditionally.  For directories that is fine.  But that
> is not correct for files.  Only executables and executable scripts
> should have the execute bit set upon them.
> 
> What you wanted to set was:
> 
>   chmod -R u+rwX,go-rwx /home/user/Documents

I ran this command to restart the process :

 find /home/user/Documents -type f -exec chmod u+rw,go-rwx -R {} \;

and will make executable all following files according the needs.


> The capital 'X' is the trick.  The GNU chmod documentation on this says:
> 
>   27.2.4 Conditional Executability
>   
> 
>   There is one more special type of symbolic permission: if you use `X'
>   instead of `x', execute/search permission is affected only if the file
>   is a directory or already had execute permission.
> 
>  For example, this mode:
> 
>a+X
> 
>   gives all users permission to search directories, or to execute files if
>   anyone could execute them before.

Yeah I did see that in the man pages but I had too much files with
hazardous rights to trust this command.


> But wait!  There's more.

Be sure I'm not going anywhere. :D


> That is usually called UPG (User Private Group).
>> chown user:user -R /home/user/Documents
> 
> And so that group should belong to the user.  Most importantly that
> group should belong *solely* to the user.  No other users should be in
> that group.  Therefore the better thing to do is to keep the group
> permissions when removing other permissions.
> 
>   chmod -R o-rwx /home/user/Documents
> 
> Then you don't need to do anything more.  That would correspond to a
> user "umask 07" setting.  better set "umask 07" or new files will be
> created with permissions you are trying to avoid.
> 
> Personally I always use "umask 02" and then only add extra protection
> to specific files and directories that I want.
> 
> And of course all of this is only important if you are operating on a
> multiuser server that has other people logging into it as non-root.
> (Root does not matter in either case.  You can't protect yourself from
> root.)  If this is on your personal laptop and no one else logs in
> then none of this matters aand I would stick with the Debian UPG
> default along with the default "umask 02".

After reading this, I actually found that :

umask and level of security : The umask command be used for setting
different security levels as follows:

umask value Security level  Effective permission (directory)
022 Permissive  755
026 Moderate751
027 Moderate750
077 Severe  700

in there :
http://www.cyberciti.biz/tips/understanding-linux-unix-umask-value-usage.html

And I was planning to set a "severe" security plan. Based on the
thinking that I have 3 computers (that only I use) to run behind a box
and that I thought wiser to set them to the maximum security first, find
out what they will exchange in second and then update the permissions
accordingly, as I have very little impact on the box security.

I then opted for the umask 077. I'm not sure if it's really justified
but it couldn't do no harm.. I guess. :)


> If you want to verify what chmod is doing the GNU chmod command has
> the -v extension.  It will echo print what it is doing while it is
> doing it.  Adding the -v would show helpful information.  For example:
> 
>   $ chmod -v -R 700 junk
>   mode of `junk' retained as 0700 (rwx--)
>   mode of `junk/junk2' retained as 0700 (rwx--)
>   mode of `junk/junk2/file1' changed to 0700 (rwx--)

I always forget to use that functionality. ^^


>> I run :
>> find /home/user/Documents ! -perm 0700
> 
> As Linux-fan correctly noted that skips files that match 0700
> exactly.  So that part is working correctly.  What didn't work was the
> chmod 700 part.  But that was good because that isn't want you want to
> do.
> 
[...]
> I believe you must have a typo somewhere.  If you double check
> everything you will find it.  However!  As I explained you do not want
> to chmod 700 all of your files recursively.  That would be bad.  So
> take it as a good miss and don't do it again.

Strangely, it seems that using symbolic mode instead of octal solved my
issue : all files are treated and I have no random results anymore.

Very thanks for your lights again, any indicators are always
welcomed. :)

-- 
“One original thought is worth a thousand mindle

Re: experimenting with dpkg: installing on a different system

[berenger . morel]

Le 20.06.2014 18:43, Sven Joachim a écrit :

On 2014-06-20 17:42 +0200, berenger.mo...@neutralite.org wrote:

First, a warning: I am basically trying to reinvent a wheel, only 
for

my pleasure and knowledge.


Good luck.  As you had already noticed, this wheel is going to have 
some

rough corners, and just using debootstrap is much easier.


Indeed. But I do this on my spare time, without any constraint or order 
from anyone, and since I consider this for pleasure, using debootstrap 
would be a little like using a cheat.
Of course, I know that I risk to generate noise here with that, but I 
think I was clear enough: I want fun and learn, not especially something 
which works out of the box.



Yesterday, I was experimenting with dpkg ( for my own fun, and to
learn things ), especially with the parameter "--root" which change
the targeted system. Or, to be more exact, to change the directories
where it tries to retrieve it's own informations ( /var/lib/dpkg ) 
and

where it tries to deploy stuff ( / ).


You can do this, but the directory you pass as "--root" argument 
better

have some standard utilities on it, since dpkg requires them.


Now you mention it, it's true that dpkg's man says it uses a chroot. 
And indeed since the new root does not have anything, it can't run a 
script... did not thought about that.



* dpkg depends on various packages ( on my testing current system ),
namely: libc6, tar, libselinux1, liblzma5, libbz2-1.0, zlib1g.


It also needs some programs from essential packages on which it does 
not
explicitly depend.  And the maintainer scripts will use those as 
well.


I did not read the debootstrap script enough to notice those tools.




When it comes to libc6, dpkg report an error of being unable to find
various scripts, like IIRC "/var/lib/dpkg/tmp.ci/preinst" ( in the
chrooted environment, so /mnt/var/lib/dpkg/tmp.ci/preinst ).


This could be because the interpreter for those scripts, typically
/bin/sh, is missing.  Note that the exec(3) family of functions 
return

ENOENT in that case which often confuses users.
Besides a shell for the maintainer scripts, dpkg needs some utilities
like rm(1) for proper operation.  It checks for those at startup, but
that does not work with the "--root" option.

Maybe this gives you some idea why debootstrap has been invented in 
the

first place.


As I said, it's mostly for playing and fun. I understand why 
debootstrap was made (automated things which are painful otherwise). 
Trying to reinvent the wheel, or to be more precise, to reproduce it's 
behavior in my case, is not a good thing and I do not do it: at work. 
But for learning and understanding, it is a nice way imho. And I take 
some fun when tinkering around egg and chicken problems.




Cheers,
   Sven


Thanks for your reply.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/751d8abc151c25a99d84d0bac08ee...@neutralite.org

chkrootkit message

[François Patte]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Bonjour,

I get this alert message (concerning lightdm) from chkrootkit

! RUID  PID TTYCMD
! root 3153 tty7   /usr/bin/X :0 -seat seat0 -auth
/var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch

What does it mean?

- -- 
François Patte
UFR de mathématiques et informatique
Laboratoire CNRS MAP5, UMR 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)1 8394 5849
http://www.math-info.univ-paris5.fr/~patte
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlOn31IACgkQdE6C2dhV2JW7sgCfX4MfZ3opNXqPaqxS0wj2IfcB
qfIAoJ1yb8TtJp4NtrE+bPh5ARrW65z8
=mNoE
-END PGP SIGNATURE-


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/53a7df52.6080...@mi.parisdescartes.fr

Re: dhclient changes IP address

[Bob Proulx]

Rainer Dorsch wrote:
> my fritz.box is a DSL router from AVM, which unfortunately does not give me 
> access to syslog.

Oh.  I didn't realize that.  The way you talked about it I thought it
was another Debian system.

> What I noticed is that 192.168.178.87 shows up without MAC address
> in the list of network devices of the fritz.box

That is almost certainly an important clue to the problem.

> I think I will try first to get support from AVM on that topic.

Good luck!  However I am pessimistic about support from companies.  In
general they like to sell you things but unless you are a notable
customer with deep pockets they tend not to want to spend time
supporting you.  Every vendor is unique however.

In general a dhcp server will always assign the same IP address to the
same client.  Everyone knows this is based upon the client ethernet
address.  But many people do not know this is also based upon the uid
the client offers.  (It might be client hostname.  I don't remember
and have no time to research the precise thing.)  If a client offers
different uids then the dhcp server is obligated to assign a different
IP address.  The idea was that a multiple booting system would have
different uids and wouldn't share the same IP address even though it
had the same ethernet hardware address.

Is it possible that your client is somehow producing a DHCP request
but using different client parameters with the different requests and
therefore receiving different IP assignments in return?  Since you are
getting different IP assignments and it must be for some reason then I
suspect this possible case.  Assuming that the dhcp server is
standards conforming.

Do you keep getting different IP assignments every day?

> If that is not successful, I will look in more detail into the
> tcpdumps (although since I have to take that on the client side,
> that might be difficult during the startup phase).

Note that dhcpdump can still be used on your client.  It would require
some patience but it would capture the exchange.  Easier than
wireshark and specific to the domain at hand.  It would be on my list
to try to capture what the differences are when getting the different
IP assignments.

Bob


signature.asc
Description: Digital signature