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

«‹›1029 »

ExcessBLarg!: Sep 1, 2001

hootimus posted:

What's going on here?

Glibc is probably resolving .local via a method other than dns lookup.

What does your /etc/host.conf, /etc/nsswitch.conf, and /etc/resolv.conf contain? Are you running avahi?

# ? Feb 26, 2011 22:47

Adbot: ADBOT LOVES YOU

# ? May 17, 2024 14:32

spiritual bypass: Feb 19, 2008; Grimey Drawer

I've set up a VirtualHost on Apache but I get 403 every time I try to connect to http://test
Is there something wrong with these config entries? I can get to the default host fine.
I've chmodded everything in projects to 777

code:

<Directory /home/rt4/projects>
	AllowOverride All
	Order Allow,Deny
	Allow from All
</Directory>

<VirtualHost *:80>
	DocumentRoot "/home/rt4/projects/test"
	ServerName test
</VirtualHost>

# ? Feb 26, 2011 22:50

Comatoast: Aug 1, 2003; by Fluffdaddy

Doesn't <Directory> need to be inside the <VirtualHost> ?

# ? Feb 27, 2011 00:00

spiritual bypass: Feb 19, 2008; Grimey Drawer

I don't think so. Putting it inside VirtualHost didn't change anything. Maybe my vhost is fine and I've done something else wrong?

edit: eh, I'm gonna give nginx a shot

spiritual bypass fucked around with this message at 01:10 on Feb 27, 2011

# ? Feb 27, 2011 00:07

pram: Jun 10, 2001

Are you using selinux, what does your apache error_log say?

Just tail -f your error_log and refresh the page.

# ? Feb 27, 2011 01:14

jromano: Sep 24, 2007

I have an old P2 Compaq with 64mb that I've been trying to install Linux on, since XP runs extremely slow. However, I can't get any of the Live CDs to boot. They never make it past the text loader. So far I've tried:

Linux Mint LXDE (wishful thinking)
Crunchbang (goes to a black screen)
Puppy (actually boots, and even got wireless working, but locks up at the installer)
Slatiz (it hangs at kernel_thread_helper)

Any ideas of how I can get this thing up and running? I'd prefer Linux over an old version of Windows.

# ? Feb 27, 2011 01:16

spiritual bypass: Feb 19, 2008; Grimey Drawer

Pram posted:

Are you using selinux, what does your apache error_log say?

Just tail -f your error_log and refresh the page.

No SELinux here. The log says:

[error] [client 127.0.0.1] (13)Permission denied: access to / denied

edit: Dammit! I had the directory permissions set right for my project directory, but not my home directory. :sweatdrop:

spiritual bypass fucked around with this message at 01:46 on Feb 27, 2011

# ? Feb 27, 2011 01:43

Bob Morales: Aug 18, 2006; ~~Just wear the fucking mask, Bob~~

I don't care how many people I probably infected with COVID-19 while refusing to wear a mask, my comfort is far more important than the health and safety of everyone around me!

jromano posted:

I have an old P2 Compaq with 64mb

You're going to need more RAM. Depending on what speed that machine is, you'll need either 66MHz or 100MHz SDRAM (350+ MHz will have the 100MHz bus). PC100 is generally backwards compatible, though. Not sure about PC133.

That said, you'll be best off with something like Slackware. Give OpenBSD or FreeBSD a try as well. OpenBSD is very light-weight. Load up FluxBox and you can even run Firefox.

# ? Feb 27, 2011 03:32

My Rhythmic Crotch: Jan 13, 2011

ExcessBLarg! posted:

Glibc is probably resolving .local via a method other than dns lookup.

What does your /etc/host.conf, /etc/nsswitch.conf, and /etc/resolv.conf contain? Are you running avahi?

You get a gold star! My nsswitch.conf was hosed up. It used to contain this:

code:

hosts: files mdns4_minimal [NOTFOUND=return] dns

And I changed it to this:

code:

hosts: files dns mdns4_minimal

Works great now, thanks!

# ? Feb 27, 2011 17:58

ColdPie: Jun 9, 2006

Is there a way to prevent a directory from being stupidly deleted, while having it act like a normal directory in every other regard? rm -rf important/ should fail, while mkdir important/stuff/ and touch important/file should both succeed. Unfortunately, I think this is impossible with standard UNIX permissions, but I feel like there must be some common workaround. Any ideas?

# ? Feb 27, 2011 20:01

ExcessBLarg!: Sep 1, 2001

hootimus posted:

My nsswitch.conf was hosed up.

I wouldn't say that's "hosed up" so much as "someone else decided that .local should only exist in the mdns universe," which is an arguably reasonable if not short-sighted policy.

Which distribution is this? It's worth knowing which are going to cause this kind of trouble in the future.

# ? Feb 27, 2011 20:23

My Rhythmic Crotch: Jan 13, 2011

ExcessBLarg! posted:

Which distribution is this? It's worth knowing which are going to cause this kind of trouble in the future.

Suse 11.2 and I've also checked on the 11.4RC2 and it has it as well

# ? Feb 27, 2011 20:40

My Rhythmic Crotch: Jan 13, 2011

ColdPie posted:

Is there a way to prevent a directory from being stupidly deleted, while having it act like a normal directory in every other regard? rm -rf important/ should fail, while mkdir important/stuff/ and touch important/file should both succeed. Unfortunately, I think this is impossible with standard UNIX permissions, but I feel like there must be some common workaround. Any ideas?

I think you want the chattr command for that

# ? Feb 27, 2011 20:40

Matt Zerella: Oct 7, 2002; Norris'es are back baby. It's good again. Awoouu (fox Howl)

ColdPie posted:

Is there a way to prevent a directory from being stupidly deleted, while having it act like a normal directory in every other regard? rm -rf important/ should fail, while mkdir important/stuff/ and touch important/file should both succeed. Unfortunately, I think this is impossible with standard UNIX permissions, but I feel like there must be some common workaround. Any ideas?

I do this on our company's FTP server. Read up on "chattr", that's what I used.

# ? Feb 27, 2011 20:43

drukqs: Oct 15, 2010; wank wank you're a pro vaper I'm not wooptiedoo...

I'm in the process of attempting to reflash a Wyse terminal and make it into a little linux box. I'm following this guide:

http://web.archive.org/web/20080407133601/http://thunderlord.net.pl/evo/

using this script:

http://web.archive.org/web/20080420192133/www.mowson.org/karl/articles/linux-netxfer/

and having problems with tftpd.

this is the error I'm getting:

"Feb 27 13:49:12 htpc tftpd[578]: unknown option -?"

I have modified the script only a little, just changed "eth0" to "eth1" but I haven't changed it in any other way.

I have made sure that tftpd and dhcpd are installed on my machine. <edit> and as you can see there is no "tftpd -?" in the script... so maybe the question mark refers to a command my version of tftpd can't do?

drukqs fucked around with this message at 03:11 on Feb 28, 2011

# ? Feb 27, 2011 23:02

crazyfish: Sep 19, 2002

blink 182 posted:

lsof?

The command returns very very quickly and only happens once a minute. It would take a near-miracle for lsof's repeat mode to catch it in the act, though I can give it a shot.

# ? Feb 28, 2011 00:47

JHVH-1: Jun 28, 2002

crazyfish posted:

The command returns very very quickly and only happens once a minute. It would take a near-miracle for lsof's repeat mode to catch it in the act, though I can give it a shot.

How about writing up a little script to output lsof to a file on access using inotifywait

That would be easier than doing it manually anyway.

# ? Feb 28, 2011 00:52

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

crazyfish posted:

Short of printk()s in the kernel, is there a way for me to be able to tell which process is issuing a particular obscure SCSI command to a particular device?

Processes shouldn't be issuing SCSI commands. That's the kernel's job, and if that's going wrong, it probably indicates an issue with the kernel/drivers or hardware rather than a user-mode program. Your first step is to trace down the syscall making that command, which can be done by stracing all running processes or by making good use of SystemTap.

Vulture Culture fucked around with this message at 04:16 on Feb 28, 2011

# ? Feb 28, 2011 04:04

crazyfish: Sep 19, 2002

Misogynist posted:

Processes shouldn't be issuing SCSI commands. That's the kernel's job, and if that's going wrong, it probably indicates an issue with the kernel/drivers or hardware rather than a user-mode program. Your first step is to trace down the syscall making that command, which can be done by stracing all running processes or by making good use of SystemTap.

Well aware of the distinction, though it's quite possible for a userland application to initiate the process (think smartd). In any case, I haven't heard of SystemTap before - I'll be sure to look into that. Thanks.

# ? Feb 28, 2011 06:24

ExcessBLarg!: Sep 1, 2001

Misogynist posted:

Processes shouldn't be issuing SCSI commands.

J�rg Schilling and libscg disagrees with this.

He wrote cdrecord & libscg (which cdrecord uses) for Solaris/Linux/whatever. The idea is kind of cool, libscg is a generic SCSI transport layer so that the SCSI bus/commands may be observed/issued by userspace programs directly. This allowed cdrecord to be ported to a bunch of Unix variants that implemented their kernel-level SCSI stacks differently. While this made his CD burning tools widely available, they were for a long time a giant pain the rear end for Linux users to use. Especially when it was insisting on making SCSI calls directly to drives that weren't even SCSI devices (ATAPI, a SCSI-speaking ATA hybrid) and so required a bunch of hackery in the kernel to actually work.

He's been mad at Debian (and Linux in general) ever since Debian folks patched cdrecord to make it a bit easier to use (be able to run as a non-root user, allow users to specify drives uses a device name instead of SCSI bus:device:lun notation, etc.) He flipped his poo poo during a license controversy a few years later, and I don't think has been heard of much since.

ExcessBLarg! fucked around with this message at 17:20 on Feb 28, 2011

# ? Feb 28, 2011 17:18

hackedaccount: Sep 28, 2009

Perhaps a tool like dtrace or SystemTap would help.

EDIT: Oops, Message-o-nist beat me to it.

# ? Feb 28, 2011 17:41

pseudorandom name: May 6, 2007

ExcessBLarg! posted:

He flipped his poo poo during a license controversy a few years later, and I don't think has been heard of much since.

Thankfully, OpenSolaris came along and took one for the team.

# ? Feb 28, 2011 20:39

pram: Jun 10, 2001

ExcessBLarg! posted:

J�rg Schilling

I never had any idea about this, the drama is kind of hilarious:

code:

Note: The author of cdrecord should not be bothered with problems in this version.

# ? Feb 28, 2011 20:44

TheOtherContraGuy: Jul 4, 2007; brave skeleton sacrifice

Is there a good Ubuntu equivalent of EndNote?

# ? Mar 1, 2011 02:39

calandryll: Apr 25, 2003; Ask me where I do my best drinking!; Pillbug

TheOtherContraGuy posted:

Is there a good Ubuntu equivalent of EndNote?

Not quite the same thing, but I like Mendeley. It is a PDF manager, with a bunch of added bonuses. It can also work as a citation manager. Zortero (sp?) is a Firefox extension that some people really like.

# ? Mar 1, 2011 02:43

uG: Apr 23, 2003; by Ralp

If I am already using nginx and want a reverse proxy, what should I use? It seems like most people use nginx as a reverse proxy in front of apache, and installing 2 versions of nginx sounds messy.

The reason I want to do this is to stop slow connections from taking up all my fast cgi sockets.

# ? Mar 1, 2011 06:45

pram: Jun 10, 2001

Squid?

# ? Mar 1, 2011 06:55

uG: Apr 23, 2003; by Ralp

Pram posted:

Squid?

That was the obvious choice, but it seems nginx is set to overtake squid in that category too. I guess I was really hoping for a way to do this with just nginx

# ? Mar 1, 2011 08:41

pram: Jun 10, 2001

You could try Apache and mod_proxy but that might defeat the purpose. :mmmhmm:

# ? Mar 1, 2011 09:38

covener: Jan 10, 2004; You know, for kids!

uG posted:

That was the obvious choice, but it seems nginx is set to overtake squid in that category too.

Does that matter?

# ? Mar 1, 2011 12:42

JHVH-1: Jun 28, 2002

uG posted:

If I am already using nginx and want a reverse proxy, what should I use? It seems like most people use nginx as a reverse proxy in front of apache, and installing 2 versions of nginx sounds messy.

The reason I want to do this is to stop slow connections from taking up all my fast cgi sockets.

Apache behind nginx works great if it suits your needs better.

Does this do what you want http://nginxlimitproxy.sourceforge.net ?

There are also some built in limiting functions if you search the nginx wiki.

# ? Mar 1, 2011 17:50

Modern Pragmatist: Aug 20, 2008

I'm the only Linux user in a mostly Windows work environment. I have several services that are used around the lab including smb, ssh, and svn. The problem is that none of the computers are able to ping my computer using my hostname. When I ping my IP address instead, it works fine. I could technically go around and put an entry in everyone's hosts file since I have a static IP, but there has to be another way.

I believe that the computers are using UDP broadcast for their hostname lookup but I'm not sure what settings I haven't toggled just right to get them to recognize my computer via hostname.

Any help would be greatly appreciated.

I am running Fedora 14

# ? Mar 1, 2011 22:30

Profane Obituary!: May 19, 2009; This Motherfucker is Dead

uG posted:

If I am already using nginx and want a reverse proxy, what should I use? It seems like most people use nginx as a reverse proxy in front of apache, and installing 2 versions of nginx sounds messy.

The reason I want to do this is to stop slow connections from taking up all my fast cgi sockets.

I never tried this, but what if you setup a server block in nginx to listen to 127.0.0.1:10000 or something, and have that pass to fcgi. Then have your normal server block for your website, proxy to 127.0.0.1:10000. I think that might work?

# ? Mar 1, 2011 22:34

bort: Mar 13, 2003

Modern Pragmatist posted:

The problem is that none of the computers are able to ping my computer using my hostname. When I ping my IP address instead, it works fine. I could technically go around and put an entry in everyone's hosts file since I have a static IP, but there has to be another way.

Is your nmbd daemon running? chkconfig --list nmb or service nmb status

That will let clients resolve your hostname using NetBIOS.

e: extra d

bort fucked around with this message at 23:04 on Mar 1, 2011

# ? Mar 1, 2011 22:56

Modern Pragmatist: Aug 20, 2008

bort posted:

Is your nmbd daemon running? chkconfig --list nmb or service nmb status

That will let clients resolve your hostname using NetBIOS.

e: extra d

Interesting. For some reason it wasn't running. When I toggle that it works. Now, to add it as a startup task I should be able to do:

code:

chkconfig --levels 3 nmb on

right? It doesn't seem to actually start it.

The other option is rc.local but that's kinda messy. Any ideas?

# ? Mar 1, 2011 23:17

pram: Jun 10, 2001

Pretty sure smb starts nmb, at least in RHEL. Use chkconfig smb on

# ? Mar 1, 2011 23:22

enotnert: Jun 10, 2005; Only women bleed

Well, netbios is normally dickish (I can never get it working right at home, something about masters fighting and blah blah blah).

Easiest way is to add manually to hosts, or find whoever controls your DNS and add it into your DNS resolution.

# ? Mar 1, 2011 23:31

Modern Pragmatist: Aug 20, 2008

Looks like everything is working just dandy for now. Thanks for the help. I'll keep my fingers crossed that it continues to function as expected.

# ? Mar 1, 2011 23:37

bort: Mar 13, 2003

Modern Pragmatist posted:

chkconfig --levels 3 nmb on

right? It doesn't seem to actually start it.

Yeah, chkconfig just sets the runlevel (I'd just use chkconfig nmb on, the defaults are usually right). That's why I put the service command there, too.

quote:

Well, netbios is normally dickish

Yes, it's a horrible protocol, but it was what he was looking for. Hosts isn't easier. :kiddo:

bort fucked around with this message at 00:43 on Mar 2, 2011

# ? Mar 2, 2011 00:39

Adbot: ADBOT LOVES YOU

# ? May 17, 2024 14:32

NOTinuyasha: Oct 17, 2006; The Great Twist

uG posted:

If I am already using nginx and want a reverse proxy, what should I use? It seems like most people use nginx as a reverse proxy in front of apache, and installing 2 versions of nginx sounds messy.

The reason I want to do this is to stop slow connections from taking up all my fast cgi sockets.

Just out of curiosity, how does this setup benefit you?

# ? Mar 2, 2011 05:37

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

«‹›1029 »