Re: how to test disk for bad sector

[tomas]

On Sat, Aug 29, 2020 at 09:03:54PM -0400, Gene Heskett wrote:

[...]

> dd is a whole device writer, but his command line is to an existing 
> partition?

No. 'dd' doesn't care. How/why should it? It can write to whatever is
supposed to be organized as a sequence of blocks. I.e. a block device
(which may sit on top of a "disk device" or a partition, but also a
regular file on your file system.

If there is one thing UNIX got pretty well compared to its siblings
of the time, then it was interface abstraction (it's got still a ton
of quirks, but you've got to look at the others aroud it to really
appreciate).

Cheers
 - t


signature.asc
Description: Digital signature

Re: Regarding libdpdk not found

[tomas]

On Sun, Aug 30, 2020 at 02:18:46PM +0800, Utkarsh Kathuria wrote:
> I am currently working on pktgen for dpdk in centos7.
> When I try to make in pktgen-dpdk directory it says dependency libdpdk not
> found.
> Please help me.

You go to a terminal and type in 'apt search libdpdk'. This is my sample
session:

  tomas@trotzki:~$ apt search libdpdk
  Sorting... Done
  Full Text Search... Done
  libdpdk-dev/stable 18.11.8-1~deb10u1 amd64
Data Plane Development Kit (basic development files)

So it eems you have to install the package libdpdk-dev (depending on your 
distro,
version numbers might vary).

Hope that helps
 - t


signature.asc
Description: Digital signature

Migrating to a new disk.

[Rick Thomas]

OK, I've got a Debian computer where the system disk is showing signs of 
flakiness.  I want to replace it with a new disk and retire the old one.

Before I do it for real, I'm doing a dry-run on a vmware virtual machine.  I 
don't *think* the fact that it's virtual should affect my results.  But I put 
it out there, just incase.

Here's what I've done so far:

*) Set up a VM with one virtual SATA disc drive, and installed Buster on it.  
The system has a MBR partitioning with /boot in /dev/sda1 and the rest of the 
disk as an LVM volume-group ("tryout-vg") partition in /dev/sda5 with root, 
swap, and home as LVs.

*) Added a virtual SATA disk drive and partitioned it the same as above -- 
/dev/sdb1 is boot and /dev/sdb5 is the LVM partition.  However, in order to 
have both disks available for mounting (see below) at the same time, the new 
drive's LVM volume-group had to have a different name ("new-vg").

*) Used rsync to copy the contents of the boot, root, and home partitions from 
the original disk to the new one.

*) Modified the /etc/fstab on the new disk to reflect the names and uuid's of 
the partitions on the new disk.

*) Booted the Buster install DVD in rescue mode and ran "reinstall boot loader" 
for the new disk.

*) Rebooted and told the BIOS to boot from the new disk.  It went to the grub 
screen and proceeded to boot.

*) To my surprise, after it booted, I logged in and saw that the root, swap, 
home and boot partitions that were mounted were all from the original disk!

So what am I missing?  How do I tell grub on the new disk to use the root 
partition and volume-group on the new disk?

Thanks for any help!
Rick

Re: Re: Download button disabled

[pierre . frenkiel]

Thank you John, that works perfectly
--
Envoyé depuis l'application myMail pour Android

Re: Re: Download button disabled

[pierre . frenkiel]

Thank you Andrei, your recipe worked perfectly, as the one proposed by John
--
Envoyé depuis l'application myMail pour Android

Re: how to test disk for bad sector

[Andy Smith]

Hello,

On Sun, Aug 30, 2020 at 12:00:38AM +, Long Wind wrote:
> top-posting is default behavior of yahoo mail. i don't know why
> it's inconvenient for some list users.

I wouldn't normally bother to mention¹ but since you seem to be
pushing back on this…

Between your top posting and the HTML mails, I find it very
difficult to read your emails so I mostly haven't bothered. It would
be nice if you would care about the people who read your emails,
just because it's nice to be nice. But if you don't care about being
nice, maybe you care more about actually getting help than you care
about what the default behaviour of yahoo mail is. Up to you.

Cheers,
Andy

¹ Seems a bit too off-topic, and mostly a waste of time trying to
  change people's minds.

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

Re: how to test disk for bad sector

[Charles Curley]

On Sun, 30 Aug 2020 14:02:48 +
Andy Smith  wrote:

> Between your top posting and the HTML mails, I find it very
> difficult to read your emails so I mostly haven't bothered.

Hear, hear. My sentiments exactly.

Yahoo mail is broken. I encourage Mr. Wind to get another mail reader.

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

Re: Migrating to a new disk.

[Joe]

On Sun, 30 Aug 2020 02:44:02 -0700
"Rick Thomas"  wrote:

> OK, I've got a Debian computer where the system disk is showing signs
> of flakiness.  I want to replace it with a new disk and retire the
> old one.
> 
> Before I do it for real, I'm doing a dry-run on a vmware virtual
> machine.  I don't *think* the fact that it's virtual should affect my
> results.  But I put it out there, just incase.
> 
> Here's what I've done so far:
> 
> *) Set up a VM with one virtual SATA disc drive, and installed Buster
> on it.  The system has a MBR partitioning with /boot in /dev/sda1 and
> the rest of the disk as an LVM volume-group ("tryout-vg") partition
> in /dev/sda5 with root, swap, and home as LVs.
> 
> *) Added a virtual SATA disk drive and partitioned it the same as
> above -- /dev/sdb1 is boot and /dev/sdb5 is the LVM partition.
> However, in order to have both disks available for mounting (see
> below) at the same time, the new drive's LVM volume-group had to have
> a different name ("new-vg").
> 
> *) Used rsync to copy the contents of the boot, root, and home
> partitions from the original disk to the new one.
> 
> *) Modified the /etc/fstab on the new disk to reflect the names and
> uuid's of the partitions on the new disk.
> 
> *) Booted the Buster install DVD in rescue mode and ran "reinstall
> boot loader" for the new disk.
> 
> *) Rebooted and told the BIOS to boot from the new disk.  It went to
> the grub screen and proceeded to boot.
> 
> *) To my surprise, after it booted, I logged in and saw that the
> root, swap, home and boot partitions that were mounted were all from
> the original disk!
> 
> So what am I missing?  How do I tell grub on the new disk to use the
> root partition and volume-group on the new disk?
> 
> Thanks for any help!

The stage you missed was to chroot into the new disk root and update
grub from there. This is most painlessly done by booting the installer
medium and choosing rescue, either graphical or otherwise. When you
reach the screen that asks you where the root of the installed OS is,
pick the new root drive, mount the efi/boot partition if you're on efi,
then chroot into the root directory. Run update-grub from there.

The alternative with just boot or non-Debian rescue media is to boot
it, mount the new root somewhere in the rescue filesystem, then chroot
to it, again then update grub. Here's one description:
https://howtoubuntu.org/how-to-repair-restore-reinstall-grub-2-with-a-ubuntu-live-cd

-- 
Joe

Re: Migrating to a new disk.

[David]

On Sun, 30 Aug 2020 at 20:03, Rick Thomas  wrote:

> So what am I missing?  How do I tell grub on the new disk to use the root 
> partition and volume-group on the new disk?

Hi, I have written this message from memory
without testing any of the commands, and
I don't actually use any of these commands myself
so I hope it doesn't contain any errors.

You wrote that you copied /boot. This means that you
copied the grub configuration file (/boot/grub/grub.cfg)
unchanged, so the new grub is configured to boot
the old system.

Even though you reinstalled the bootloader on the
new drive, you kept the old config file telling it to
to boot the old drive.

One answer might be to run 'update-grub' but I can't
give you exact instructions because I don't like that
aspect of grub so I prefer to write my own grub.cfg
files without all the bloat.

I expect doing that would use os-prober to generate
a new grub.cfg file with a menu that offers to boot
any of all the operating systems that os-prober finds.
I think that's the usual default method.

I think you could backup /boot/grub/grub.cfg and just try it
and see what happens. If you backup to for example
/boot/grub/grub.cfg-BACKUP then if something goes wrong
you can just enter
'configfile grub.cfg-BACKUP' at the grub> prompt to use
the old version.

Why can't I move the document root for a site in Apache 2?

[Gary Dale]

I'm running Apache 2.4.38-3+deb10u3 on a Debian/Stable server on an 
AMD64 machine.


When I create a virtual host under /var/www, everything works as 
expected. However, if I change the virtual host's document root to 
another folder on the same machine, I get


|Forbidden You don't have permission to access this resource. 
Apache/2.4.38 (Debian) Server at .local Port 80 |


where I use .local instead of the live site's actual TLD to refer to my 
local server.


I get the same thing if I replace the public_html folder with a link to 
the other folder. To be clear, the folder and files in it are owned by 
the same account & group. And I can cd to the other folder through the 
link, so it's working.


Also to be clear, when I go to .local with the site in 
/var/www/.local/public_html, it works.


A reason I want to move the sites is that /var is in my system 
partition, which runs off of a small SSD, while the other folder in on a 
RAID-6 array with lots of space.


When I search for the problem, I see a lot of "solutions" that say just 
change the document root of the vhost and restart Apache 2. However that 
isn't working in my case. This is likely Debian specific but all the 
Debian stuff only shows vhosts under /var/www, which isn't what I want.


Any ideas?

Top-posting (was Re: how to test disk for bad sector)

[Charlie Gibbs]

On Sun, 30 Aug 2020 16:30:01 +0200
Charles Curley  wrote:

> On Sun, 30 Aug 2020 14:02:48 +
> Andy Smith  wrote:
>
>> Between your top posting and the HTML mails, I find it very
>> difficult to read your emails so I mostly haven't bothered.
>
> Hear, hear. My sentiments exactly.
>
> Yahoo mail is broken. I encourage Mr. Wind to get another mail reader.

If someone can't be bothered to take the time to write a readable
message, I can't be bothered to take the time to decipher it.

As for Outlook, I've been told that the correct pronunciation is
"Look out!"

--
/~\  Charlie Gibbs  |  Microsoft is a dictatorship.
\ /|  Apple is a cult.
 X   I'm really at ac.dekanfrus |  Linux is anarchy.
/ \  if you read it the right way.  |  Pick your poison.

Re: non-smart debian phone

[Dan Hitt]

On Sat, Aug 29, 2020 at 9:07 PM Keith Bainbridge 
wrote:

> On 28/8/20 6:23 am, Dan Hitt wrote:
> >
> > So next time i search, i hope to do it more intelligently.
> >
>
> So why not start the search now, while you have time to think about it
> for a bit before you need to jump in?
>
>
Why, indeed?

I suppose that would entail getting a list of carriers and a list of
phones, and then checking the phones against the carriers.

There seem to be sites that do compatibility checks (e.g.,
willmyphonework.net) so i guess this is possible in principle, although i
could not find either 'librem' or 'purism'  on their list.  But they do
have a way of submitting queries for items not on their list, so maybe i'm
not trying hard enough.  (Although i am looking for something more modest
than librem.)

Anyhow, i'm all ears to hear of anybody's experience in choosing a phone
which runs free software (like debian, or some of the others that people
have mentioned) and what carrier they use.  Especially if it is an
out-of-the-box experience, but also if it involves installation on an
existing phone.

And, if you install a new os on an existing phone, do you first have to
activate the phone with some carrier, or can you just plug your phone into
a monitor and keyboard and do the installation of the os just like you
would on your pc, and then go through the activation with a carrier?

And thanks for posting, Keith!

dan

Re: Why can't I move the document root for a site in Apache 2?

[john doe]

On 8/30/2020 6:27 PM, Gary Dale wrote:

I'm running Apache 2.4.38-3+deb10u3 on a Debian/Stable server on an
AMD64 machine.

When I create a virtual host under /var/www, everything works as
expected. However, if I change the virtual host's document root to
another folder on the same machine, I get

|Forbidden You don't have permission to access this resource.
Apache/2.4.38 (Debian) Server at .local Port 80 |

where I use .local instead of the live site's actual TLD to refer to my
local server.

I get the same thing if I replace the public_html folder with a link to
the other folder. To be clear, the folder and files in it are owned by
the same account & group. And I can cd to the other folder through the
link, so it's working.

Also to be clear, when I go to .local with the site in
/var/www/.local/public_html, it works.

A reason I want to move the sites is that /var is in my system
partition, which runs off of a small SSD, while the other folder in on a
RAID-6 array with lots of space.

When I search for the problem, I see a lot of "solutions" that say just
change the document root of the vhost and restart Apache 2. However that
isn't working in my case. This is likely Debian specific but all the
Debian stuff only shows vhosts under /var/www, which isn't what I want.

Any ideas?




What are the permissions of the directory in question?

Do you have a directory directive for that location in apache2?

--
John Doe

Re: Why can't I move the document root for a site in Apache 2?

[john doe]

On 8/30/2020 7:08 PM, john doe wrote:

On 8/30/2020 6:27 PM, Gary Dale wrote:

I'm running Apache 2.4.38-3+deb10u3 on a Debian/Stable server on an
AMD64 machine.

When I create a virtual host under /var/www, everything works as
expected. However, if I change the virtual host's document root to
another folder on the same machine, I get

|Forbidden You don't have permission to access this resource.
Apache/2.4.38 (Debian) Server at .local Port 80 |

where I use .local instead of the live site's actual TLD to refer to my
local server.

I get the same thing if I replace the public_html folder with a link to
the other folder. To be clear, the folder and files in it are owned by
the same account & group. And I can cd to the other folder through the
link, so it's working.

Also to be clear, when I go to .local with the site in
/var/www/.local/public_html, it works.

A reason I want to move the sites is that /var is in my system
partition, which runs off of a small SSD, while the other folder in on a
RAID-6 array with lots of space.

When I search for the problem, I see a lot of "solutions" that say just
change the document root of the vhost and restart Apache 2. However that
isn't working in my case. This is likely Debian specific but all the
Debian stuff only shows vhosts under /var/www, which isn't what I want.

Any ideas?




What are the permissions of the directory in question?

Do you have a directory directive for that location in apache2?

--
John Doe



That is, if you change the 'DocumentRoot' directive you also need to
modify or add a corresponding directory directive in apache2.
Look at '/etc/apache2/apache2.conf' for how it is done for '/var/www'.:

"
Options Indexes FollowSymLinks
AllowOverride All
Require all granted
"

I would change the path of '

Re: Migrating to a new disk.

[Andrew Cater]

The easiest method might be to boot from installation media, run the
"rescue" install, chroot to the disk you want to use, then run update-grub
from there

All the very best, as ever,

Andy C.

On Sun, Aug 30, 2020 at 2:44 PM David  wrote:

> On Sun, 30 Aug 2020 at 20:03, Rick Thomas  wrote:
>
> > So what am I missing?  How do I tell grub on the new disk to use the
> root partition and volume-group on the new disk?
>
> Hi, I have written this message from memory
> without testing any of the commands, and
> I don't actually use any of these commands myself
> so I hope it doesn't contain any errors.
>
> You wrote that you copied /boot. This means that you
> copied the grub configuration file (/boot/grub/grub.cfg)
> unchanged, so the new grub is configured to boot
> the old system.
>
> Even though you reinstalled the bootloader on the
> new drive, you kept the old config file telling it to
> to boot the old drive.
>
> One answer might be to run 'update-grub' but I can't
> give you exact instructions because I don't like that
> aspect of grub so I prefer to write my own grub.cfg
> files without all the bloat.
>
> I expect doing that would use os-prober to generate
> a new grub.cfg file with a menu that offers to boot
> any of all the operating systems that os-prober finds.
> I think that's the usual default method.
>
> I think you could backup /boot/grub/grub.cfg and just try it
> and see what happens. If you backup to for example
> /boot/grub/grub.cfg-BACKUP then if something goes wrong
> you can just enter
> 'configfile grub.cfg-BACKUP' at the grub> prompt to use
> the old version.
>
>

Re: Top-posting

[Felix Miata]

Charlie Gibbs composed on 2020-08-30 09:14 (UTC-0700):

> On Sun, 30 Aug 2020 16:30:01 +0200 Charles Curley wrote: 
>> On Sun, 30 Aug 2020 14:02:48 + Andy Smith wrote:

>>> Between your top posting and the HTML mails, I find it very
>>> difficult to read your emails so I mostly haven't bothered.

>> Hear, hear. My sentiments exactly.

>> Yahoo mail is broken. I encourage Mr. Wind to get another mail reader. 

> If someone can't be bothered to take the time to write a readable
> message, I can't be bothered to take the time to decipher it.

> As for Outlook, I've been told that the correct pronunciation is
> "Look out!"

I pronounce it out house.

"Even the magic 8 ball has an opinion on email clients: Outlook not so good"
-- 
Evolution as taught in public schools, like religion,
is based on faith, not on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: how to test disk for bad sector

[Marco Möller]

On 30.08.20 03:03, Gene Heskett wrote:

On Saturday 29 August 2020 20:39:29 Marco Möller wrote:


On 30.08.20 00:01, Long Wind wrote:
(...)


[liveuser@localhost-live  ~]$ sudo
dd if=/dev/zero of=/dev/sda1 bs=8M status=progress


(...)


14386462720 bytes (14 GB, 13 GiB) copied, 151.001 s, 95.3 MB/s
dd: error writing '/dev/sda1': No space left on device
1719+0 records in
1718+0 records out
14418915840 bytes (14 GB, 13 GiB) copied, 153.319 s, 94.0 MB/s


Unfortunately, this looks bad for your disk, because you said in
another post that your disk is supposed to provide a storage space of
320 GB. But in this test only 14 GB could be written to it and then it
is claimed to already have been filled completely. So, in context of
another post of yours, where you say that you could install a system
to a new disk mounted to that same computer, you found the problem:
the supposed to be 320 GB hard disk is either broken in a very strange
way, or it is even a fake disk sold to you as a fraud.

Wishing you good luck with your new disk and new installation!
Marco.


Uh, I have to ask. Or maybe I came into the middle of what has become a
boring thread?

dd is a whole device writer, but his command line is to an existing
partition? His command line should not have had the 1 according to my
thinking.  His whole disk might be 320 gigabytes as /dev/sda, but how
big is sda1?

Cheers, Gene Heskett



You are right, Gene!  I did not notice that it was written to partition 
"/dev/sda1", because I didn't expect this test to only be written to a 
partition instead of test writing to the whole disk. I have not noticed 
this detail in the answer when replying to it, shame on me.

My recommendation was to use the command like this:
  sudo dd if=/dev/zero of=/dev/WritingToThisDisk bs=8M status=progress
Long Wind, the test works only if you overwrite the full disk. So, as 
Gene pointed out, after the test which you did by now it cannot be 
concluded what I concluded in my former answer!


We by now only know that writing to the 14 GB of partition "/dev/sda1" 
worked correctly, but we do not know where on the disk this 14 GB are 
actually located and we also still do not know about the rest of the 
disk and if it continues to work fine also for a bigger data load 
occupying the electronics for longer time.


So, if you are interested to still know about your old drive, although 
you said that you already installed successfully to a new drive, then 
you may want to direct the test all the device and not only a partition 
of it.


Best regards,
Marco.

Re: Migrating to a new disk.

[David Christensen]

On 2020-08-30 02:44, Rick Thomas wrote:

OK, I've got a Debian computer where the system disk is showing signs of 
flakiness.  I want to replace it with a new disk and retire the old one.


I would get an SSD of the same or larger size and use dd(1) to copy the 
raw blocks of the old drive to the new drive.



David

Re: Top-posting (was Re: how to test disk for bad sector)

[Charles Curley]

On Sun, 30 Aug 2020 09:14:16 -0700
Charlie Gibbs  wrote:

> If someone can't be bothered to take the time to write a readable
> message, I can't be bothered to take the time to decipher it.

On the other tentacle, this sort of thing is usually the province of
newbies. I think it would help to refer newbies to some advice. I would
refer them to ESR's How To Ask Questions The Smart Way, but the
catb.org server is not co-operating.

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

Re: how to test disk for bad sector

[Charles Curley]

On Sun, 30 Aug 2020 22:02:21 + (UTC)
Long Wind  wrote:

> i don't have choice. gmail is blocked in china. i've tried some free
> chinese mail provider, they block debian list. (i've sent
> subscription request thru their mail service, then nothing happened)

Not good.

Proton mail? https://protonmail.com/ Anyone got any other suggestions?

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

Re: how to test disk for bad sector

[Long Wind]

 

On Sunday, August 30, 2020, 10:06:55 PM EDT, Charles Curley 
 wrote:  
 
Not good.

Proton mail? https://protonmail.com/ Anyone got any other suggestions?


Thank Charles,  Proton mail isn't blocked here now. 

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

  

Re: make a drawing

[Russell L. Harris]

On Mon, Aug 31, 2020 at 05:06:20AM +0200, Emanuel Berg wrote:

Anyway, I'd like to make a drawing of it. Any idea
what software might be used?


sweethome
inkscape
dia

Re: make a drawing

[Doug McGarrett]

On 8/30/20 11:51 PM, Russell L. Harris wrote:

On Mon, Aug 31, 2020 at 05:06:20AM +0200, Emanuel Berg wrote:

Anyway, I'd like to make a drawing of it. Any idea
what software might be used?


sweethome
inkscape
dia

There are a number of Linux-usable CAD programs, but the one I thought 
was the best is now gone: DraftSight. There is supposed to be a cloud 
version of it, but I don't know any more about that. It used AutoCAD 
commands, which I used to be familiar with (25 years ago!)  If you have 
any cad experience, try and find an app that uses the commands you know, 
or used to. If you just want a PICTURE, the best  advice is to just take 
one-or two, or however many different views you need.


--doug