The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›2 »

kyuss: Nov 6, 2004

I spent the last three days installing Ubuntu Lite and setting up a video recording machine for my dad, and so far it has been a great experience. Linux support on the internet seems to have greatly improved in the last few years.

However I have a problem:

I want to use mencoder to encode analog TV from a Hauppauge WinTV PCI FM card. Watching TV with tvtime, xawtv or mplayer works flawlessly, but I'm not getting any sound when I try to record into a file:

code:

sudo mencoder -tv driver=v4l:width=320:height=240:norm=PAL:channel=E7:chanlist=europe-west:outfmt=yuy2:audioid=0
-ovc lavc -lavcopts vcodec=mpeg4:vbitrate=900 -oac mp3lame -lameopts cbr:br=64 -o output-mp4-audio0.avi tv://

When I open the file in Windows Media Player (on another machine) it tells me:

video codec: ffdshow MPEG-4 Video Decoder
audio codec: -

The internet told me to make sure that the recording device should be set to "line" (as that is the input the TV card is connected with), but no dice.

So far I'm unsure if this is a mixer problem or a codec problem. Any ideas?

kyuss fucked around with this message at 08:18 on May 29, 2007

# ¿ May 29, 2007 08:15

Adbot: ADBOT LOVES YOU

# ¿ Apr 18, 2024 17:06

kyuss: Nov 6, 2004

It works!

I was about to post the whole output of mencoder, when I saw

code:

=================================================================
 WARNING: YOU ARE USING V4L DEMUXER WITH V4L2 DRIVERS!!!
 As the V4L1 compatibility layer is broken, this may not work.
 If you encounter any problems, use driver=v4l2 instead.
 Bugreports on driver=v4l with v4l2 drivers will be ignored.
=================================================================

So I changed the command line accordingly to

code:

oliver@ubuntu:~/video-test$ sudo mencoder -tv
driver=v4l2:width=320:height=240:norm=PAL-BG:channel=E7:chanlist=europe-west:outfmt=yuy2
-ovc lavc -oac lavc -lavcopts vcodec=mpeg4:vbitrate=900:acodec=mp3:abitrate=32 -o output-mp4-audio0.avi tv://

and I have both sound and video now. Apparently the old driver version didn't even try to include the sound part in the .avi file.

Seems all I needed was a final push in the right direction - thanks teapot

# ¿ May 29, 2007 15:24

kyuss: Nov 6, 2004

dfn_doe posted:

Linux works great with vmware as the host and/or vm. I'm not sure what problems you had before, but Linux is solidly the most stable OS for running on top of or underneath a virtualization system.

I have three different distributions installed currently, and it works like a charm. The only thing that initially bugged me was that the integrated VMware desktop lags quite a bit. Installing tightvncserver on the linux machine and accessing it by TightVNC instead of the VMware interface helped a lot here - same is for Putty for easy, SSH encrypted shell access

# ¿ May 29, 2007 19:30

kyuss: Nov 6, 2004

TheWevel posted:

I can not for the life of me get my new monitor to display its native resolution in Ubuntu. I have edited the xorg.conf file to add the native res, and it always comes back at 1280x1024 at 75hz. The native resolution is 1400x1050 at I assume 60hz. The manual doesn't state what the refresh rate should be. Manually editing the "VertRefresh" line to say "50-60" doesn't do anything... X restarts at 1280x1024 at 75hz.

Just one page ago some guy had a very similar problem. Have you checked that out?

# ¿ Jun 1, 2007 17:36

kyuss: Nov 6, 2004

I'm rebuilding my own LiveCD of Ubuntu with the help of UCK, using the standard Ubuntu LiveCD as a base for my own.

In order to free up some space, I removed stuff like Gimp and OpenOffice, however, my CD Image doesn't seem to become smaller :confused:

I guess these packages must either still hide in some cache directory, or they leave their traces in the actual filesystem itself. Any ideas?

# ¿ Jun 10, 2007 07:28

kyuss: Nov 6, 2004

teapot posted:

No, they are huge, and I had to remove them when building custom CD with reconstructor. Of course, you have to remove all openoffice-related packages, not just openoffice package itself.

But when you remove them with the package manager, they should be gone from your image, leaving you with more free space, right?

I'm still having trouble with the packet managing concept. How would I remove OpenOffice and Gimp with apt-get or aptitude, for example? Am I really expected to manually select every single lib / plugin / spellchecking module? And sometimes, when I choose to remove a certain package, completely unrelated stuff like "ubuntu-desktop" are automatically selected for removal, too.

There must be an easier way to remove software, right? :confused:

I just want to get rid of some stuff to be able to burn my iso to a CD.

# ¿ Jun 10, 2007 09:18

kyuss: Nov 6, 2004

Twlight posted:

I�m working during my internship and I�m focusing on linux administration. While I am doing other projects the position leaves me with a lot of free time to build different things with some of the old hardware. (Pentium 3�s). I�ve built an IDS as well as written some perl scripts but I�m fresh out of ideas for a new side project. Anyone have any ideas or would like to share what they�re working on?
Thanks

My next project will be a linux-based multimedia server that will record stuff from TV, hold and play all my other movies, and store backups from my other PC. Setting up a VWware environment for testing purposes would be cool as well

# ¿ Jun 11, 2007 18:48

kyuss: Nov 6, 2004

Alright, it's me again with a modified Ubuntu LiveCD (Desktop, x86 architecture) that frequently crashes on my Dad's PC

Basically what I did was take the original Ubuntu Desktop LiveCD, free up some space by removing OpenOffice and Gimp, and add programs like sshd, ntfs-config, tvtime, mplayer and mencoder to it. The idea was to have a LiveCD my Dad could boot from to record some old VHS tapes with a TV card his lovely XP64 doesn't support.

Testing the CD on VMware showed no problems, so I burned it on CD and tried it out on the real PC: first it seemed to work, but then it began to crash with increasing frequency. First it was every ten minutes or so, at the end I could barely open a shell or click on a menu entry without Ubuntu terminally crapping itself

I suspected some thermal issue first, but the CD drive doesn't get particularly hot :confused:

Could it be the fact that my Dad's machine has an Athlon 64 CPU?
Is there some error log I could check for unusual entries?
Maybe running everything from RAM instead of the CD would help - how would I do this?

Any help is, as always, appreciated

# ¿ Jun 14, 2007 06:57

kyuss: Nov 6, 2004

teapot posted:

0. Post your hardware configuration. How would anyone be able to spot a known-unreliable hardware or something that is not supposed to work until some proprietary driver is installed?

Update: the culprit was some ancient ISDN card we were totally oblivious of, and that was hogging precious IRQs. The ensuing gangrape of the last free IRQ by graphics, sound and TV card was apparently causing the lockups

# ¿ Jun 17, 2007 18:02

kyuss: Nov 6, 2004

Quick question:

I just installed SlugOS on the Linksys NSLU2 Device
I want lighttpd+mod_webdav to give me access to my files from everywhere
For access control, I need to use htpasswd/htaccess
As the apache2-utils aren't available for some reason (and I don't need the full apache package), I installed mini_httpd from here instead

My problem is: htpasswd seems to be installed, but I cannot run it for some reason:

code:

root@openslug:/usr/sbin# ls -l
total 364
lrwxrwxrwx  1 root root     17 Apr 22 21:49 brctl -> ../../bin/busybox
lrwxrwxrwx  1 root root     17 Apr 22 21:49 chpasswd -> ../../bin/busybox
lrwxrwxrwx  1 root root     16 Apr 22 20:25 chroot -> chroot.coreutils
-rwxr-xr-x  1 root root  12404 Mar  2 18:39 chroot.coreutils
lrwxrwxrwx  1 root root     17 Apr 22 21:49 crond -> ../../bin/busybox
-rwxr-xr-x  1 root root   5807 Aug  5  2007 htpasswd
lrwxrwxrwx  1 root root     17 Apr 22 21:49 httpd -> ../../bin/busybox
lrwxrwxrwx  1 root root     17 Apr 22 21:49 loadfont -> ../../bin/busybox
lrwxrwxrwx  1 root root     17 Apr 22 21:49 rdate -> ../../bin/busybox
-rwxr-xr-x  1 root root  14320 Mar  2 16:39 readprofile
-rwxr-xr-x  1 root root  10768 Jan  1  1970 rngd
lrwxrwxrwx  1 root root     17 Apr 22 21:49 setlogcons -> ../../bin/busybox
-rwxr-xr-x  1 root root 263192 Jan  1  1970 sshd
lrwxrwxrwx  1 root root     17 Apr 22 21:49 telnetd -> ../../bin/busybox
-rwxr-xr-x  1 root root   5804 Mar  2 16:39 tunelp
-rwxr-xr-x  1 root root   6311 Jan  1  1970 update-inetd
-rwxr-xr-x  1 root root   4131 Jan  1  1970 update-modules
-rwxr-xr-x  1 root root  16308 Jan  1  1970 update-passwd
-rwxr-xr-x  1 root root   4409 Jan  1  1970 update-rc.d
root@openslug:/usr/sbin#

I tried everything I remember from linux:

htpasswd
./htpasswd
sudo, su...

to no avail. The file cannot be found somehow. Example:

code:

root@openslug:/usr/sbin# ./htpasswd
-sh: ./htpasswd: not found
root@openslug:/usr/sbin#

Any ideas?

# ¿ Apr 25, 2009 17:57

kyuss: Nov 6, 2004

kyuss posted:

Quick question:

I just installed SlugOS on the Linksys NSLU2 Device

I want lighttpd+mod_webdav to give me access to my files from everywhere

For access control, I need to use htpasswd/htaccess

As the apache2-utils aren't available for some reason (and I don't need the full apache package), I installed mini_httpd from here instead

My problem is: htpasswd seems to be installed, but I cannot run it for some reason:

After virtually every package I tried to install needed serious tweaking, I said gently caress this. Installing Debian now.

# ¿ Apr 27, 2009 20:32

kyuss: Nov 6, 2004

Does anyone have experience with setting up Dovecot+Postfix?

I am able to access Dovecot via IMAP now, but after changing from Mailbox to Maildir format, Dovecot doesn't seem to be able to pick up mails anymore that have been

a) created locally by mutt / mail
b) picked up by Postfix (SMTP)

I'm kinda stumped as to which logfiles to look in / how this MTA/LDA stuff really works.

# ¿ May 5, 2009 03:53

kyuss: Nov 6, 2004

falz posted:

You probably need to adjust your mail_location in dovecot.conf to tell it where your maildir lives relative to your user's homedir. It's incredibly flexible and it can be placed just about anywhere, including not under your user's home dir at all. Perhaps the most reasonable place for it for you is wherever mutt thinks it should be.

I'll try that, thanks!

quote:

I would also highly recommend using dovecot's LDA for postfix mailbox delivery. This will generally keep things simpler and will let you use sieve for filtering/vacation replies/etc.

That's my intention too. I think my problem is that I cobbled several installation Howtos for postfix / dovecot / deliver together, resulting in an inconsistent install.

Postal posted:

You should be able to install NDISWrapper in a Live CD, but it won't stick between reboots. Google and NDISWrapper HOWTO for more information.

There's sufficient info on creating your own custom Ubuntu Live CD on the web. I tried this last year and it was a breeze.

# ¿ May 7, 2009 06:09

kyuss: Nov 6, 2004

Hi,

I recently took the final step and completely switched over from Windows to Linux on my home machines. Over time I grew quite fond of Microsoft Powershell and its object-oriented approach to things: dealing with files means dealing with objects that have properties like .size, .fullname, .extension etc.

What are my options under Linux here? I'll go back to parsing strings if I have to, but I'd rather not. On a quick look Python's 'os' module seemed quite rudimentary.

# ¿ Jan 12, 2011 08:04

kyuss: Nov 6, 2004

kyuss posted:

Hi,

I recently took the final step and completely switched over from Windows to Linux on my home machines. Over time I grew quite fond of Microsoft Powershell and its object-oriented approach to things: dealing with files means dealing with objects that have properties like .size, .fullname, .extension etc.

What are my options under Linux here? I'll go back to parsing strings if I have to, but I'd rather not. On a quick look Python's 'os' module seemed quite rudimentary.

Anyone?

Apart from this, anyone using one of these?

My NSLU2+Debian home server seems to have kicked the bucket, and I'm looking for a worthy successor. A low power, preferrably silent home server that runs on linux and is powerful enough for providing SSH, OpenVPN and NFS.

# ¿ Jan 17, 2011 20:00

kyuss: Nov 6, 2004

waffle iron posted:

Sheevaplug

Just gave in and bought the eSATA variant from their UK reseller

Didn't get a timely answer from http://www.ionicsplug.com/ so far, and their german reseller wasn't any better. Didn't fall for the Sheevaplug's successor either, as the Guruplug appears to have terrible overheating problems.

My NSLU2 box seems to be stuck in a permanent rebooting cycle despite doing every Redboot / upslug2 procedure I could find.

So its a new box then.

# ¿ Jan 20, 2011 01:57

kyuss: Nov 6, 2004

Bash question:

The idea is to check regularly for disk activity of a certain drive. If it appears to be idle, the drive receives a spindown command.

My problem is that the string compare between $state and $newstate always comes out as not equal, even if the contents of both variables are the same.

What am I doing wrong?

code:

#!/bin/sh

disk=$1

state=`cat /tmp/mydiskstats`
newstate=`grep " $disk " /proc/diskstats`

if [ "$state" = "$newstate" ]; then
        #this is never met
        echo "drive is idling"; echo state: $state; echo newstate: $newstate

else
        echo "disk activity detected"; echo state: $state; echo newstate: $newstate
fi

echo $newstate > /tmp/mydiskstats

# ¿ Jan 30, 2011 12:10

kyuss: Nov 6, 2004

Thanks for the advice guys, putting my $newline variable in quotes did the trick.

FISHMANPET posted:

Try replacing it with '==' or '-eq'.

I tried these too, but '-eq' failed because the contents aren't numbers only, and '==' is considered an "illegal operator" (no idea why, using standard bash here).

covener posted:

I'd also recommend using cp and diff instead of cat and string comparisons.

Misogynist posted:

Still won't be equal -- echo adds a newline. You want echo -n "$newstate" >/tmp/mydiskstats.

Good calls, will remember.

BagelMaster posted:

I'm assuming it's because I haven't actually logged in and started up a desktop session, so there's nothing to show, thus rendering VNC useless. Is there an easy way around this?

Have you tried neatX? I'll prefer it over VNC any time.

# ¿ Jan 31, 2011 21:52

kyuss: Nov 6, 2004

iOS question:

a buddy of mine jailbroke his iPhone and wonders how to restart a daemon through SSH. Both "service" and "init" don't seem to be available. Even "ps" doesn't seem to exist, not sure about "kill". At least he claims to be root.

Any advice?

# ¿ Feb 25, 2011 07:19

kyuss: Nov 6, 2004

Sizzlechest posted:

I'm trying to set up a better backup system for my Synology file server. The backup application that comes with it seems to just do a sync rather than an incremental backup.

I'm not familiar with Linux backup software/methods. I was considering using git to handle the incremental backup part and just rsync to an external drive every week. It's just data files, so I don't care about capturing the file attributes or backing up SQL data.

Try the "duplicity" backup system with "duply" as its frontend, it's awesome. It's an official debian package too, so it's super easy to install.

# ¿ Mar 11, 2011 16:55

kyuss: Nov 6, 2004

Lee Van Queef posted:

Is there a way to backup iPhone apps and app data in Linux?

Never tried it myself, but supposedly there is.

First, install the "libimobiledevice-utils" package for basic iPhone support.

Then, use "ideviceinstaller -a" from "git clone http://git.sukimashita.com/ideviceinstaller.git"
to store your apps into a zip file which you can copy elsewhere for backup.

# ¿ Apr 17, 2011 09:31

kyuss: Nov 6, 2004

After updating to Ubuntu 11.04 hibernating fails with a "PM: Not enough free memory" message.

Free RAM is 800/2048MB, free disk space is ~15GB. Any ideas how to solve this?

# ¿ May 3, 2011 17:02

kyuss: Nov 6, 2004

Quick routing question:

My routing table:

code:

Kernel-IP-Routentabelle
Ziel            Router          Genmask         Flags   MSS Fenster irtt Iface
192.168.2.0     0.0.0.0         255.255.255.0   U         0 0          0 wlan0
10.1.106.0      0.0.0.0         255.255.255.0   U         0 0          0 eth0
169.254.0.0     0.0.0.0         255.255.0.0     U         0 0          0 eth0
0.0.0.0         192.168.2.254   0.0.0.0         UG        0 0          0 wlan0
0.0.0.0         10.1.106.1      0.0.0.0         UG        0 0          0 eth0

I am unable to ping an adress like "10.1.2.1" as long as the "192.168.2.254" default route exists. After deleting it, everything's dandy until the next reboot.

Question is: how do I fix this the proper way? Would this help?

My /etc/network/interfaces is

code:

auto lo
iface lo inet loopback

allow-hotplug wlan0
iface wlan0 inet dhcp
         wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf

allow-hotplug eth0
iface eth0 inet static
address 10.1.106.100
netmask 255.255.255.0
gateway 10.1.106.1

kyuss fucked around with this message at 18:29 on May 31, 2011

# ¿ May 31, 2011 18:20

kyuss: Nov 6, 2004

Bob Morales posted:

You can't have two 'default' routes, only one can be the default.

That's what I thought. But this:

code:

0.0.0.0         192.168.2.254   0.0.0.0         UG        0 0          0 wlan0
0.0.0.0         10.1.106.1      0.0.0.0         UG        0 0          0 eth0

shows that there are two default routes at the same time, or am I mistaken here?

# ¿ May 31, 2011 18:41

kyuss: Nov 6, 2004

Will put my route in rc.local, thanks guys

For the record, I have a ton of switches in a lab setup that I'm tasked to run some tests on.

I got tired from permanently switching chairs, so I put a notebook in the testing setup that sits both in the testing network (10.x.x.x) and the company WLAN (192.168.x.x).

When connecting from my work notebook via the company WLAN to the testing notebook, I have the aforementioned routing table on the testing machine.

I'm still curious about the multiple default route mechanics, though.

# ¿ Jun 1, 2011 08:01

kyuss: Nov 6, 2004

I was pissed off by Unity, so I decided to switch distributions.

I spent 89� on a SSD and decided to play around a bit:

Fedora 15 - picked it over its Gnome3 desktop, liked it at first, but then it started to crash on me, especially after waking up from standby / hibernate. Installed the catalyst driver for my graphics card on a hunch, but this made things worse - got garbled graphics in menus. I felt much of this was due to Fedora being licensing nazis, so I decided leave it at that.

Linux Mint 11 - nice GUI, but it felt old after trying out Gnome3 for a few days. Tried to install Gnome3, had the GUI permanently crashing on me right after login, decided to try another distro.

Arch Linux - ugly text installer that kept giving me poo poo because I tried to change the partition layout. Had to manually pick a console font because I dared to change the keyboard layout to other than US - weird. Installation was very quick after that. First fuckup was at boot, Grub somehow got my partition layout wrong. Fixed it by hand. Second fuckup after changing to Grub2 - it had the wrong settings from Grub1. Fixed it again.

Installing xserver and Gnome3 on the other hand was a breeze. Had some trouble initially getting sound in flashplugin, because I didn't read its installation instructions carefully enough.

So far Arch feels lightning fast, and easily configurable. Hardware support seems great. Currently, I have to start Gnome3 manually via "startx", but I kinda like it, will propably leave it at that.

# ¿ Jun 3, 2011 14:09

kyuss: Nov 6, 2004

Guys, guys...

http://linux.die.net/man/1/fdupes

# ¿ Jun 28, 2011 06:50

kyuss: Nov 6, 2004

Question 1: as a happy user of "screen", I often find myself working at runlevel 3 for a while, and then switching to runlevel 5 now and then. However, I'm unable to reconnect to a screen I started at runlevel 3 while in xterm on runlevel 5. Any ideas how to do this?

Question 2: I have a beefy DB2 (linux) server at work that performs abysmally and I may be tasked to fix this some day. It's a setup from last year, with 16 SAS disks configured as RAID6 and ample RAM. However, it's response times are considerably slower than the old system it it supposed to replace, with virtually no load on it.

So this reeks of an elementary config problem to me, however I know next to nothing about db2 and database servers. Any input where I may start in resolving this mess?

Thank you.

# ¿ Jul 10, 2011 15:17

kyuss: Nov 6, 2004

bort posted:

screen -list and then connect to the particular session, e.g. screen -r pts-0

Well I'll be damned, it just works. Must have hosed up somewhere else along the way for having problems with it.

telcoM posted:

For a serious analysis, more information would be good. What is the type/model of the RAID controller? Is it a real hardware-accelerated RAID controller, or is RAID6 implemented at the driver level, and the hardware is just a "basic" SAS controller?

If it is a real hardware RAID controller, does it include a write cache unit?

What's the access pattern of your application like? In other words, what is the use of the database like?

How is the disk space allocated? You said you have 16 disks - are they configured as one big RAID6 set, or as two or three sets according to the purpose: one set for data, another for archive logs, and maybe a third for indexes.

Thank you so much for your input, I'll deliver as soon as I get to it.

hackedaccount posted:

use iotop or just top. is the writer process at 100% utilization or is waiting on I/O often?

Those will be my first steps. Thank you guys

# ¿ Jul 11, 2011 11:17

kyuss: Nov 6, 2004

Question: I have a new external USB HDD hooked up to a SheevaPlug. The HDD isn't auto mounted at boot, despite the appropriate settings in etc/fstab.

code:

# /etc/fstab: static file system information.
#
# Use 'vol_id --uuid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name devices
# that works even if disks are added and removed. See fstab(5).
#
# <file system> <mount point>   <type>  <options>       <dump>  <pass>
proc            /proc           proc    defaults        0       0
UUID=bcc2ce72-afdb-4980-b86f-89e347a9b249 /               ext3    errors=remount-ro 0       1
UUID=e16add4f-446e-492a-af57-9648ec1eab69 /boot           ext2    defaults        0       2
UUID=d2140c10-1b1e-4f64-90a0-f5b987a58f30 none            swap    sw              0       0
UUID=902df9d4-fcf4-4f7a-85e5-59a2d0e8ef7c /media/usb0 auto defaults,rw,user,auto,exec 0 0

I tried another USB HDD, and it is auto-mounted fine. What could be going on here?

# ¿ Jul 14, 2011 20:58

kyuss: Nov 6, 2004

Yes it has, and it is mounting fine from a terminal, as in "mount /media/usb0". It's just that it's not auto-mounted after a reboot. Makes me wonder if this is a timing problem, maybe the mount process happens before the USB drive is available?

# ¿ Jul 14, 2011 21:51

kyuss: Nov 6, 2004

Sweet, i'll try the udev way. Thanks

# ¿ Jul 15, 2011 06:54

kyuss: Nov 6, 2004

code:

#!/bin/sh
find ./Musik/ -name *.mp3 -execdir mp3gain -r {} \; < ohnoes.txt

Background info: "mp3gain" sometimes throws a dialogue that can be answered "y/n" style. As the answer is always "no" and I want to run the script unattended, I feed several hundred lines containing a singular "n" from the "ohnoes.txt" file.

Weird problem: the script works perfectly in a debian bash shell, but it doesn't work on my PC that's running Arch. On debian, every question is automatically answered "n", while on Arch I run into an endless loop of the same dialogue.

What's the deal here?

vvvvvvv

code:

yes n | find ./Musik/ -iname '*.mp3' -execdir mp3gain -p -r {} \+;

That did the trick actually. Thanks!

kyuss fucked around with this message at 15:54 on Jul 23, 2011

# ¿ Jul 23, 2011 09:06

kyuss: Nov 6, 2004

My next project is to digitize all my paperwork of the last years, and put it all in one big digital archive. I'll propably use the document scanner at work to make PDFs, and run an OCR tool on it afterwards.

I want to use PDF metadata to their full advantage, but I'm stumped at how to use them efficiently from an outside (file browser?) perspective.

Is there some indexing tool that runs under linux and is metadata-aware? Or should I try the non-minimalist approach and use some sort of open source document managing system? As always, any help is appreciated

# ¿ Jul 30, 2011 09:02

kyuss: Nov 6, 2004

kyuss posted:

My next project is to digitize all my paperwork of the last years, and put it all in one big digital archive. I'll propably use the document scanner at work to make PDFs, and run an OCR tool on it afterwards.

I want to use PDF metadata to their full advantage, but I'm stumped at how to use them efficiently from an outside (file browser?) perspective.

Is there some indexing tool that runs under linux and is metadata-aware? Or should I try the non-minimalist approach and use some sort of open source document managing system? As always, any help is appreciated

Current status:

I intend to create some sample PDFs @300dpi on monday, and try my luck with tesseract + hocr2pdf. This should eventually give me sandwich PDFs, with the original scan as image overlaid by the OCR results as invisible text. Bonus points for creating PDF/A. I'll let Google Desktop index these PDFs afterwards for full text indexing, and stick to meaningful folder names, as that's propably all I need.

# ¿ Jul 30, 2011 21:55

kyuss: Nov 6, 2004

spankmeister posted:

Take a look at Cacti.

It's a webbased monitoring tool. It can graph all sorts of things like disk usage, network speed, temperature, fan speed etc etc...

I'm currently on a project involving Cacti, and I know the learning curve can be steep. Feel free to ask in this thread if you have any questions.

# ¿ Sep 22, 2011 17:08

kyuss: Nov 6, 2004

taqueso posted:

Pretty much. It is a dangerous command for a superuser on anything but a personal machine.

Speaking of which, yesterday I did an innocent "sudo service --status-all" on a production machine, and got a face full of services trying to start up (instead of the expected listing of the service status).

Any ideas how this happens? The machine in question is on Debian Lenny, so maybe the "service" command changed a bit over time?

# ¿ Oct 13, 2011 06:58

kyuss: Nov 6, 2004

nitrogen posted:

Nope, it SHOULD do what you want.

Were you getting soemthing like this?
code:
 sudo service --status-all
 [ + ]  acpid
 [ - ]  bootlogd
 [ - ]  bootlogs
 [ ? ]  bootmisc.sh
 [ ? ]  checkfs.sh

...

Nah, thats what I expected to get, but the output was something different. As if every service in this list had received a startup command, regardless if it was already running or not.

Didn't help that the head of development was sitting right next to me as I tried to get a basic overview over his production server

Oh well.

# ¿ Oct 15, 2011 08:02

kyuss: Nov 6, 2004

Zom Aur posted:

Looking at the sources, my guess is either the BFQ I/O scheduler or unreliable hardware. I'm guessing the former, since I've encountered similar issues with the BFQ scheduler.

You can switch back to CFQ easily by just booting with elevator=cfq on your kernel line in grub/lilo/whatever. Try it and see if the problem persists.

The concept of different I/O schedulers was completely new to me, so I looked around a bit and found in Aaron Carroll's I/O scheduling on RAID (scroll down a bit) a really good paper on this topic.
I'm really learning a lot from this thread, thank you guys.

# ¿ Nov 15, 2011 21:08

Adbot: ADBOT LOVES YOU

# ¿ Apr 18, 2024 17:06

kyuss: Nov 6, 2004

Social Animal posted:

Is there an easy way to access a Windows share in Linux? I don't know if share is the right word actually. Basically I have a Windows box and I want it to share files with my Linux box but I'm not sure how to go about this.

You can always mount your shares the classic way, as either manually from the terminal, or an appropriate entry in /etc/fstab.

Typing from memory, a terminal entry could look like this:

"mkdir /mnt/mywindowsshare && mount -t cifs \\windowshost\someshare /mnt/mywindowsshare -o user=windowsuser%passsword "

Look it up with "man mount.cifs" or "man fstab".

fletcher posted:

What commands should I run on a remote server to ensure that the next tenant doesn't try to get creative with data recovery tools?

You may want to start with clearing the local history ("history -c") in case you entered any sensitive information (cleartext user / password combos).

Shaocaholica posted:

Neato.

Seems like you can set a different IO scheduler per disk:

http://www.cyberciti.biz/faq/linux-change-io-scheduler-for-harddisk/

Additional info I found thanks to this very thread:

http://google-opensource.blogspot.com/2008/08/linux-disk-scheduler-benchmarking.html

http://www.ertos.nicta.com.au/publications/#ug ("I/O scheduling on RAID")

# ¿ Jan 7, 2012 12:17

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›2 »