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

«‹›1041 »

Megaman: May 8, 2004; I didn't read the thread BUT...

Misogynist posted:

If you put your custom facts into a Puppet module, they'll be automatically distributed to your clients every time they connect to the master, though I assume from your previous post that you've already gotten this far.

You don't actually need this information in Facter (but it can be useful if you want to MCollective query against it later). If you want to do things the "Puppet way," you should write an External Node Classifier that looks up the provided instance ID via the EC2 API and assigns applied classes based on the direct results of the DescribeInstances API call. You could do the same thing by matching a catch-all class against a default node and doing what you're doing with the tags, though, and it will work fine. Your approach will actually be a little bit cleaner if you're dealing with both EC2 and hosts somewhere else (physical, another cloud provider, etc.).

I've read about node classifiers, and I've used a fact script and mysqlDB for cfengine a long time ago. But to tell you the truth, I actually can't see why I'd use one now. I'm using just another .pp file to take the user-data information, assign classes to them, and assign modules to those classes.

I guess the concern now is: should I use a file on the machine to define multiple facts about class, for instance a box could have this AND that class based on geolocation, etc., or should I just stick with user-data. My feeling is that a file on the machine is much more flexible, my concern is that user-data in aws is either ONLY 1 word, or ONLY 1 script, but not both, and cannot also be a string of attributes as I mentioned slightly above. I wouldn't want to forfeit user-data if I ever needed it in the future for something else and have to rearchitect the way I gather node class information. Is this a reasonable concern?

# ? Sep 23, 2013 02:35

Adbot: ADBOT LOVES YOU

# ? Jun 12, 2024 22:03

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

Megaman posted:

I've read about node classifiers, and I've used a fact script and mysqlDB for cfengine a long time ago. But to tell you the truth, I actually can't see why I'd use one now. I'm using just another .pp file to take the user-data information, assign classes to them, and assign modules to those classes.

I guess the concern now is: should I use a file on the machine to define multiple facts about class, for instance a box could have this AND that class based on geolocation, etc., or should I just stick with user-data. My feeling is that a file on the machine is much more flexible, my concern is that user-data in aws is either ONLY 1 word, or ONLY 1 script, but not both, and cannot also be a string of attributes as I mentioned slightly above. I wouldn't want to forfeit user-data if I ever needed it in the future for something else and have to rearchitect the way I gather node class information. Is this a reasonable concern?

95%+ of people are using EC2 user data for the sole purpose of feeding cloud-init. Don't reinvent the wheel here.

User data can contain whatever you want, though, and the "ONLY 1 script" piece is completely inaccurate. cloud-init works on multipart MIME archives containing whatever data you feel like stuffing into them, though working with an archive format is going to be complicated from the Puppet side. I'd probably use a role tag or something similar, and leave my user data well enough alone.

# ? Sep 23, 2013 02:50

evol262: Nov 30, 2010; #!/usr/bin/perl

BlackMK4 posted:

Yes, I managed to get the thing working but it's still not broadcasting like prior installs. Weird. as. Hell. The other machines are OSX, which had cupsctl BrowseProtocols='"cups dnssd"' run on them.

Avahi not installed?

# ? Sep 23, 2013 16:20

nitrogen: May 21, 2004; Oh, what's a 217°C difference between friends?

Anyone know some good resources to find a good linux laptop?

I know of a few vendors, like System76, and ZaReason, but I know nothing about hardware vendors other than apple because i've been locked into OSX for the last 8 years or so.

Anyone have a few tips, or a few resources I can check out so I can be a better informed consumer?

Otherwise I will just endup getting something like a Latitude E6320, that's the laptop I have with work, and it seems to work well, albeit it's old.

# ? Sep 23, 2013 19:33

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Don't buy from ZaReason. They have no idea what they're doing. I don't know about System76.

My ThinkPad works quite well, but I've heard Lenovo is driving that brand into the ground.

# ? Sep 23, 2013 19:34

evol262: Nov 30, 2010; #!/usr/bin/perl

nitrogen posted:

Anyone know some good resources to find a good linux laptop?

I know of a few vendors, like System76, and ZaReason, but I know nothing about hardware vendors other than apple because i've been locked into OSX for the last 8 years or so.

Anyone have a few tips, or a few resources I can check out so I can be a better informed consumer?

Otherwise I will just endup getting something like a Latitude E6320, that's the laptop I have with work, and it seems to work well, albeit it's old.

There's actually nothing wrong with Apple laptops running Linux either, other than the Broadcom wireless.

# ? Sep 23, 2013 20:03

BlackMK4: Aug 23, 2006; wat.; Megamarm

evol262 posted:

Avahi not installed?

I compared packages between the non-GUI setup and a GUI install... Avahi was one of the differences and I figured that may be it - it was not. I got it working by giving OSX the print server address and a path to the printer spool below. It still doesn't broadcast the printer share. Who knows.

# ? Sep 23, 2013 21:57

evol262: Nov 30, 2010; #!/usr/bin/perl

BlackMK4 posted:

I compared packages between the non-GUI setup and a GUI install... Avahi was one of the differences and I figured that may be it - it was not. I got it working by giving OSX the print server address and a path to the printer spool below. It still doesn't broadcast the printer share. Who knows.

I'm pretty sure OSX doesnt' use cups native browsing. Does it show up in avahi-discover? is 5353 open?

# ? Sep 23, 2013 22:08

Oakland Martini: Feb 14, 2008; D&D: HASBARA SQUAD
THE APARTHEID ACADEMIC

It's important that institutions never take a stance like "genocide is bad". Now get out there and crack some of my students' skulls.

I've got a Dell T5600 workstation with dual Xeon E5-2620 processors and 4x8 GB RAM coming in the mail. I am planning on installing Ubuntu for scientific computing stuff: some Matlab, lots of C and Fortran programs. The workstation comes with Windows 7, and I am planning on dual-booting (maybe getting rid of Windows entirely at some point, but I have to use Word with MathType to write papers with my geezer colleagues who refuse to learn latex).

I've set up dual-boot with Windows and Ubuntu on various laptops and gaming PCs, but I just wanted to check this thread to see if there's anything I should be aware of when trying this with a workstation. Should I use the server version? I will work both directly and remotely.

# ? Sep 25, 2013 22:14

Precambrian Video Games: Aug 19, 2002

I've had single-cpu workstations with Centos/Red Hat and it sucks. Packages get out of date, python, perl, MPI libraries, etc. It's just a hassle and not worth it for a desktop. Having said that, most serious computing nodes/supercomputers use Redhat/Centos/some other server-oriented distro, but then usually someone's being paid to compile different versions of packages if needed.

Oh, and I don't know about the latest versions of Ubuntu/Fedora, but I never managed to get Fedora to boot from an md raid partition.

# ? Sep 25, 2013 22:22

Oakland Martini: Feb 14, 2008; D&D: HASBARA SQUAD
THE APARTHEID ACADEMIC

It's important that institutions never take a stance like "genocide is bad". Now get out there and crack some of my students' skulls.

In grad school I had access to servers with similar (perhaps more dated) specs and they were running Ubuntu. This is my personal machine and I don't have anyone to administer it for me, so I figured it would be a safe bet, especially since I have used it in other contexts.

No RAID. I don't do much data-intensive stuff so disk performance isn't an issue, and I keep all my work files on Dropbox so I don't worry about redundancy much.

# ? Sep 25, 2013 22:34

spankmeister: Jun 15, 2008

Dropbox eh? I guess you could always ask the NSA for a copy if you need a backup. :haw:

# ? Sep 25, 2013 22:39

spankmeister: Jun 15, 2008

Yeah Ubuntu is fine. Rhel/centos is really nice but not for cutting edge packages. Fedora works well too. Don't just use rhel/centos just because it's more "enterprisey".

e: apologies for the doublepost, brainfart.

# ? Sep 25, 2013 22:43

Oakland Martini: Feb 14, 2008; D&D: HASBARA SQUAD
THE APARTHEID ACADEMIC

It's important that institutions never take a stance like "genocide is bad". Now get out there and crack some of my students' skulls.

spankmeister posted:

Dropbox eh? I guess you could always ask the NSA for a copy if you need a backup.

One more reason not to worry about redundancy!

Should I go for the server version or is the desktop version fine (I guess it probably has a bunch of poo poo I don't need, but that's true for my other computers as well)?

# ? Sep 25, 2013 22:52

evol262: Nov 30, 2010; #!/usr/bin/perl

eXXon posted:

Oh, and I don't know about the latest versions of Ubuntu/Fedora, but I never managed to get Fedora to boot from an md raid partition.

Something's drastically wrong with that. GRUB2 will do it, but the old metadata=0.99/install bootloader on both disks way that every distro did for /boot followed by an initrd with mdraid drivers or a kernel with mdraid built in worked on Fedora the same way as every other distro.

# ? Sep 25, 2013 23:00

Illusive Fuck Man: Jul 5, 2004; RIP John McCain feel better xoxo ð ð; Taco Defender

Maybe not the right thread for this, but I dunno what is. I'm a cs dude who doesn't know poo poo about actually managing servers, and my job now requires me to manage servers. I have a whole bunch of servers with 1TB disks in them, and my boss wants them all to boot from the network and use the disk as temporary space or whatever.

/share/rootdir is exported by the nfs server as read only
The tftp server has pxelinux.0, pxelinux.cfg/default, kernel, and initrd (created with dracut)
The dhcp server hands out the path for pxelinux.0 and a rootpath=nfs:nfs_server:/share/rootdir
rootdir has a copy of rootdir/var in rootdir/varbkp
rootdir/etc/rc.d/rc.sysinit is modified to partition the disk (if necessary), mount /home and /var on the disk, and copy /varbkp to /var.

All of this works.

I guess this is kinda nice because I can just add another mac address to the group on the dhcp server every time I want to add another server, but I dunno. Is there a standard way to do this kinda thing? Also, at some point each server is going to need a different set of config files in /etc for some poo poo. I could have the sysinit script ssh somewhere and grab them but this seems dumb. This whole thing seems dumb. What the gently caress am I doing?

# ? Sep 25, 2013 23:12

Longinus00: Dec 29, 2005; Ur-Quan

Oakland Martini posted:

One more reason not to worry about redundancy!

Should I go for the server version or is the desktop version fine (I guess it probably has a bunch of poo poo I don't need, but that's true for my other computers as well)?

You don't need the server version if you're running it as a workstation.

# ? Sep 26, 2013 00:15

evol262: Nov 30, 2010; #!/usr/bin/perl

Illusive gently caress Man posted:

Maybe not the right thread for this, but I dunno what is. I'm a cs dude who doesn't know poo poo about actually managing servers, and my job now requires me to manage servers.

Red flags are going up.

Illusive gently caress Man posted:

I have a whole bunch of servers with 1TB disks in them, and my boss wants them all to boot from the network and use the disk as temporary space or whatever.

Up higher. Why does he want this? NFSroot is fine for thin clients, LSTP, kiosks, and stateless servers. These won't be in the future.

Illusive gently caress Man posted:

/share/rootdir is exported by the nfs server as read only
The tftp server has pxelinux.0, pxelinux.cfg/default, kernel, and initrd (created with dracut)
The dhcp server hands out the path for pxelinux.0 and a rootpath=nfs:nfs_server:/share/rootdir
rootdir has a copy of rootdir/var in rootdir/varbkp
rootdir/etc/rc.d/rc.sysinit is modified to partition the disk (if necessary), mount /home and /var on the disk, and copy /varbkp to /var.

The readonly root stuff will help you immensely here. Please look into it. You don't need rc.sysinit. Why do you even need /home? Why can't /var be tmpfs?

If it were me doing this for no reason, I'd partition the disks into a Ceph or Gluster node and mount filesystems off there so you get resiliency, performance, and some use out of all that wasted disk. Again though, why?

Illusive gently caress Man posted:

I guess this is kinda nice because I can just add another mac address to the group on the dhcp server every time I want to add another server, but I dunno. Is there a standard way to do this kinda thing? Also, at some point each server is going to need a different set of config files in /etc for some poo poo. I could have the sysinit script ssh somewhere and grab them but this seems dumb. This whole thing seems dumb. What the gently caress am I doing?

Here's where there are problems. If you need different config files in /etc, you have a few options:

Keep a copy of /etc/ on the disk somewhere and bind mount it (same readonly-root stuff).
Leave the servers stateless and provision on boot with Ansible, Puppet, Chef, or whatever. This is probably the best solution.

# ? Sep 26, 2013 00:16

SurgicalOntologist: Jun 17, 2004

I'm trying to set up guacamole so I can use my office workstation (Ubuntu 13.04) from whereever. Guac is compatible with VNC and RDP. I'm trying to decide which way to go but the terminology goes over my head. I'm reading the relevant section of the manual (http://guac-dev.org/doc/gug/configuring-guacamole.html#vnc) but I'm a bit confused by statements like:

"If you are okay with having a desktop that can only be accessed via VNC, one of these [RealVNC, TigerVNC] is likely your best choice."

and

"If you need to share your local desktop, we recommend using x11vnc rather vino, as it has proven more performant and feature-complete in our testing. If you don't need to share a local desktop but simply need an environment you can access remotely, using a VNC server like RealVNC, TigerVNC, or TightVNC is a better choice."

Does this mean:
a) with RealVNC and TigerVNC, the desktop is now ONLY a remote desktop, and cannot be accessed except remotely.
b) with RealVNC and TigerVNC, the remote user spawns a new desktop environment and so cannot interact with whatever was left running by a local user.
c) with RealVNC and TigerVNC, when a remote user is connected a local user is locked out

I'm guessing (b) because the other interpretations seem absurd but I'm not sure. If I just want to get home and keep working with whatever windows I left open at work, does that mean I should use x11vnc?

And finally, if I'm correct about the above, where does RDP fit in? Is it like x11vnc, or like the others? Any suggestion for which I should use?

Ok- one more question. In addition to serving guacamole my workstation is also serving RStudio, iPython, and Matlab servers. In my "research" I keep coming across reverse proxies. I've read up a bit and I think I understand what they do, but I can't fathom if it would be useful to me. Is it something I should consider?

# ? Sep 26, 2013 00:26

Precambrian Video Games: Aug 19, 2002

evol262 posted:

Something's drastically wrong with that. GRUB2 will do it, but the old metadata=0.99/install bootloader on both disks way that every distro did for /boot followed by an initrd with mdraid drivers or a kernel with mdraid built in worked on Fedora the same way as every other distro.

I should also say I didn't try very hard to figure out why it wasn't working - it wasn't really worth the effort.

# ? Sep 26, 2013 03:04

Computer Serf: May 14, 2005; Buglord

Any advice / recommendations / best practices for setting up a private server in one office that can be remotely accessed by employees (non-techie people) without robots or jerks messing around with the server?

This will be my first time doing something this serious, and well I "know OF" most of the tools and practices that go into this sort of thing, I have zero experience actually working with any of it.
Any guides, help, or pointing in the right direction would be greatly appreciated!

Computer Serf fucked around with this message at 06:49 on Sep 26, 2013

# ? Sep 26, 2013 06:40

hazzlebarth: May 13, 2013

Panda Time posted:

Any advice / recommendations / best practices for setting up a private server in one office that can be remotely accessed by employees (non-techie people) without robots or jerks messing around with the server?

Will this be just a file server or do you want them to have shell/desktop access as well?

# ? Sep 26, 2013 10:31

evol262: Nov 30, 2010; #!/usr/bin/perl

VNC can be used for a, b, and c, depending on how you configure it. So can NX. RDP operates similarly. Which one are you trying to do?

# ? Sep 26, 2013 16:31

SurgicalOntologist: Jun 17, 2004

I'd like to do (b). I have my workstation set up to automatically log in, there's only ever one session so I'd just like to control that session remotely.

I think I got it working with guacamole > xrdp > vino, so far I've just tested it from here (man it's trippy running a remote desktop in a browser tab on the same computer you're sitting at) but it should work remotely. If there's an easier (faster) way, I'm all ears though.

Also I'm not sure about security, I've got the guacamole login page set up with SSL (and a strong password) but then no password for the rdp layer. Should be fine, right, if I only expose the guacamole port and not the xrdp or vino ports?

E: Also running guacamole > ssh but not otherwise exposing the ssh port.

# ? Sep 26, 2013 16:42

Mr. Clark2: Sep 17, 2003; Rocco sez: Oh man, what a bummer. Woof.

For reasons outside my control I need to have the notification emails that our Opsview system (it's a Nagios fork) sends out come from a particular email address (joe_schmoe@example.com). The system is running Ubuntu and using postfix as the MTA.

I have the relay server set up properly in opsview, I can see that the emails are hitting that server but are then going to the badmail folder.

I edited the main.cf file for postfix and changed "myorigin" to joe_schmoe@example.com but when I check the relay server I can see that the emails are coming from "nagios@joe_schmoe"@.example.com so I've obviously goofed something up (I'm a linux novice).

How can I change this from address? Thanks in advance!

# ? Sep 26, 2013 19:23

fourwood: Sep 9, 2001; Damn I'll bring them to their knees.

I have a Linux HTPC that I use to play emulators/ROMs, and the screen keeps blanking after 30 minutes. I have the emulators (mednafen and ZSNES, in particular) launching out of an Openbox session, and apparently using just an Xbox 360 controller isn't sufficient to interrupt the idle timer. The only thing I've managed to Google up to fix this is putting some 'xset' directives in my ~/.config/openbox/autostart.sh file. But this doesn't seem to do the trick (although I think it changed it from 10 minutes...).

Here's what I'm setting (kitchen sink approach at the moment):

code:

/usr/bin/xset s 18000 18000
/usr/bin/xset dpms 18000 18000 18000
/usr/bin/xset -dpms
/usr/bin/xset s off
/usr/bin/xset s noblank

Am I doing something wrong here? 18000 seconds is 5 hours, but I've tried it with smaller values with no effect (3600, 7200, and 9000). Or is there another way to stop the screen blanking in Openbox?

Edit: To clarify, it's turning off the display (TV reports no signal), not just blacking out the screen.

fourwood fucked around with this message at 00:34 on Sep 27, 2013

# ? Sep 26, 2013 20:26

Delta-Wye: Sep 29, 2005

So I rebooted my workstation and a bunch of my bash scripts broke. Digging into it, sourcing a bash script now adds the current path variable to the command.

I have a script that is just:

code:

echo "# = $#, 1 = $1"

When I run it directly (./test) I get the following output:

code:

# = 0, 1 =

When I source it (. ./test), I get:

code:

# = 1, 1 = PATH=/usr/lib/lightdm/lightdm:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/opt/local/bin

So damned weird. Any ideas? This isn't normal behavior and I'm not sure if I changed something or what. I've fixed the scripts that broke (they had optional arguments that I never used anyways) so I'm good, but I'd like to figure out what the hell is going on.

# ? Sep 27, 2013 01:06

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Thinking about adding another drive (2TB) to my Windows/Ubuntu dual-boot system, so I'm thinking about the best way to share it between the operating systems. Last time I looked (long ago), linux + NTFS = pain.

Is that still the case?

Would I be better off with a couple of partitions (I'd rather not, but whatever)?

# ? Sep 27, 2013 02:09

Delta-Wye: Sep 29, 2005

Thermopyle posted:

Thinking about adding another drive (2TB) to my Windows/Ubuntu dual-boot system, so I'm thinking about the best way to share it between the operating systems. Last time I looked (long ago), linux + NTFS = pain.

Is that still the case?

Would I be better off with a couple of partitions (I'd rather not, but whatever)?

The userland ntfs drivers are perfectly fine stability-wise. I'm not sure if there is a performance hit from running your I/O through fuse or whatever, but if you're just streaming media and stuff it should be fine. (I don't know about the performance hit because I've never noticed and/or cared :shh:)

# ? Sep 27, 2013 02:17

nitrogen: May 21, 2004; Oh, what's a 217°C difference between friends?

Mr. Clark2 posted:

For reasons outside my control I need to have the notification emails that our Opsview system (it's a Nagios fork) sends out come from a particular email address (joe_schmoe@example.com). The system is running Ubuntu and using postfix as the MTA.

I have the relay server set up properly in opsview, I can see that the emails are hitting that server but are then going to the badmail folder.

I edited the main.cf file for postfix and changed "myorigin" to joe_schmoe@example.com but when I check the relay server I can see that the emails are coming from "nagios@joe_schmoe"@.example.com so I've obviously goofed something up (I'm a linux novice).

How can I change this from address? Thanks in advance!

myorigin is just for domains, and should be example.com.

http://www.postfix.org/ADDRESS_REWRITING_README.html <---read this instead.

# ? Sep 27, 2013 02:32

Longinus00: Dec 29, 2005; Ur-Quan

Thermopyle posted:

Thinking about adding another drive (2TB) to my Windows/Ubuntu dual-boot system, so I'm thinking about the best way to share it between the operating systems. Last time I looked (long ago), linux + NTFS = pain.

Is that still the case?

Would I be better off with a couple of partitions (I'd rather not, but whatever)?

I haven't had a problem with NTFS and linux in a long time. Just make sure that windows has actually unmounted the filesystem before you try to use it in linux (e.g. you can't access the drive from linux if you dual booted after hibernating or maybe some stuff related to fastboot).

# ? Sep 27, 2013 04:48

sm00th: Jan 5, 2013

fourwood posted:

code:
/usr/bin/xset s 18000 18000
/usr/bin/xset dpms 18000 18000 18000
/usr/bin/xset -dpms
/usr/bin/xset s off
/usr/bin/xset s noblank
Am I doing something wrong here? 18000 seconds is 5 hours, but I've tried it with smaller values with no effect (3600, 7200, and 9000). Or is there another way to stop the screen blanking in Openbox?

As far as I remember I only used "-dpms" and "s off".
I'm not sure but you might be reenabling screensaving with "s noblank" after "s off" even though the manpage says it only sets the preference

fourwood posted:

Or is there another way to stop the screen blanking in Openbox?

It is not Openbox who blanks the screen, its X server.

fourwood posted:

Edit: To clarify, it's turning off the display (TV reports no signal), not just blacking out the screen.

Sounds like DPMS. If it still does that after all the xset witchery it may not be applied at all, you can check your current settings with "xset q"

# ? Sep 27, 2013 09:03

Toasticle: Jul 18, 2003; Hay guys, out this Rape

Need a recommendation for a guy who hasn't worked a linux box for a couple years.

Company I work for has (set up by the guy before me) a windows 2008 box acting as the companies email, website, file server, and firewall with the DHCP duties handled by a 5 year old linksys wireless router which has been making me insane, we have 4 people who travel alot and connect to their office PCs via windows remote desktop and the port forwarding on the router is a loving nightmare to keep working properly. Also runs the Pervasive database server for our accounting software. Email is the free version of mailenable and the free version of MXScan for virus/spam filtering. Website is literally just 5 HTML pages of "Hi this is what we do" so nothing at all complicated. Lately the email/spam/virus scanning has been flaking out (Jammed up queues that wont clear without a complete shutdown/restart or hogging 50% of the systems CPU) which is probably because we get about 1500 emails a day 1400 of which are spam/viruses or his last attempt at updating mailenable screwed something up which for the life of me I cant figure out. I'd like to move everything but the file sharing/database off onto a linux box but its been 6 or 7 years since I last managed one. Was an Ubuntu (like version 8 or 9) using postfix/dovecot to handle mostly email blasts to about 18,000 subscribers for industrial machinery auctions and a fairly simple website.

Since I haven't touched linux since then was looking for suggestions for a distro to be a mail, simple website, DHCP server and firewall. I was going to just go for the same setup as I used before but am looking for any suggestions if there is distro or email software that would handle things better than Ubuntu/Postfix. Also my boss would like to see if there's something similar to an exchange server so people don't have to remote desktop in when they just need to find an email they sent 3 weeks ago from their office PCs.

# ? Sep 27, 2013 15:14

evol262: Nov 30, 2010; #!/usr/bin/perl

Toasticle posted:

Since I haven't touched linux since then was looking for suggestions for a distro to be a mail, simple website, DHCP server and firewall. I was going to just go for the same setup as I used before but am looking for any suggestions if there is distro or email software that would handle things better than Ubuntu/Postfix. Also my boss would like to see if there's something similar to an exchange server so people don't have to remote desktop in when they just need to find an email they sent 3 weeks ago from their office PCs.

Use a router for DHCP and firewall. Pay for a terminal server license and make them RDP to that, if they really need to -- don't port forward back to their PCs.

Put your website and mail on Google Apps. It's not worth the trouble to handle this yourself.

# ? Sep 27, 2013 18:01

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

There's no magic software raid thing that works across Windows and Linux is there?

As a continuance to my last question about NTFS, I found that I have two 2TB drives, and as I'm not going to have anything not backed up on them, I was thinking about using RAID-0, but I'd also like to use them across Windows and Linux.

I'm guessing I'm going to have to use hardware RAID, but just thought I'd ask.

# ? Sep 27, 2013 18:49

Toasticle: Jul 18, 2003; Hay guys, out this Rape

Unfortunately not my calls, he wants the email server in house for whatever reason and he wants RDC as do the designers so they can access their own PC's with their illustrator set up just the way they like (Clothing design company). The website is so simple I can just leave it on the windows server to be honest. I don't know or care why he wants things set up this way, I just do what I'm told.

At the bare minimum though I need to get the email off the windows server, mailenable free is poo poo and is bogging everything down never mind the constant need for restarts when it locks up.

Imagine a guy with a degree in electrical engineering (Why he went into clothing design is beyond me) so he knows enough to know whats technically possible but not enough to do it himself. He also has ADHD or something because once he gets something in his head changing his mind is not even worth the effort. On the plus side he does understand when I tell him that doing X has the potential to cause A/B/C issues and if one of them happens he accepts it ws his decision to take the risk.

Arguing with him over this stuff isn't worth the risk of starting drinking again so I just do what he wants since he's fine accepting the consequences when something goes wrong.

Toasticle fucked around with this message at 19:17 on Sep 27, 2013

# ? Sep 27, 2013 19:03

Feral Integral: Jun 6, 2006; YOSPOS

Hey you all, question about some video card and xen poo poo:

I've compiled Xen 4.4 from the git master successfully, and it seems everything is working right. However, when I boot up my system it seems like Debian is loading my radeon 5450 (the card I want to passthrough to my guest os - this is a pci x1 card!) *before* my nvidia gforce 9600 (a pci x16 card)!

Of course, I want my Dom0 (Debian) to be using the nvidia by default and leaving the Radeon available for a guest Windows 7 system. How can a specify that one card is used over the other on boot?

# ? Sep 27, 2013 19:29

evol262: Nov 30, 2010; #!/usr/bin/perl

Feral Integral posted:

Hey you all, question about some video card and xen poo poo:

I've compiled Xen 4.4 from the git master successfully, and it seems everything is working right. However, when I boot up my system it seems like Debian is loading my radeon 5450 (the card I want to passthrough to my guest os - this is a pci x1 card!) *before* my nvidia gforce 9600 (a pci x16 card)!

Of course, I want my Dom0 (Debian) to be using the nvidia by default and leaving the Radeon available for a guest Windows 7 system. How can a specify that one card is used over the other on boot?

There are a few ways to do this. Please read the stuff on xen-pciback.

Toasticle posted:

Unfortunately not my calls, he wants the email server in house for whatever reason and he wants RDC as do the designers so they can access their own PC's with their illustrator set up just the way they like (Clothing design company). The website is so simple I can just leave it on the windows server to be honest. I don't know or care why he wants things set up this way, I just do what I'm told.

This is more of a security problem. You don't have a dedicated server guy and you're running (presumably) unmaintained Windows servers for something that'd cost you $15/yr.

Email in-house is stupid for the same reason. SPF, DKIM, not being an open relay, and everything else you need to get reliable mail transport sounds like it's beyond the skillset of anyone at your company. Paying someone to deal with all the hassle is virtually free and removes a lot of headache from your very small company. There is zero reason to have email in-house.

If they want access to "their own PCs with Illustrator set up just the way they like" they should get laptops. Or hit the terminal server from inside the company as well and set up illustrator on VDI or whatever. Port forwarding to internal workstations so they can RDP in from wherever and use Illustrator "jsut the way they like" is wrong is many, many ways.

Toasticle posted:

Imagine a guy with a degree in electrical engineering (Why he went into clothing design is beyond me) so he knows enough to know whats technically possible but not enough to do it himself. He also has ADHD or something because once he gets something in his head changing his mind is not even worth the effort. On the plus side he does understand when I tell him that doing X has the potential to cause A/B/C issues and if one of them happens he accepts it ws his decision to take the risk.

It's also your job (assuming you're the "computer guy" at this shop) to inform him of what his limitations are and what you're not willing to do. When A/B/C happens and your mail server gets on a blacklist so you can't talk to clients at all anymore until you migrate your email infrastructure somewhere else, or whatever web frontent you set up for email has a zero day and you get your email server rooted, or someone brute forces the terrible password the clothing designer's set on their account which has port-forwarded RDP with no real security and steals all your clothing designs from your NAS or whatever...

It's better to stop problems preemptively than fight fires when they spring up, and they will spring up in your environment. You don't have the in-house expertise to manage email, you probably don't need to host your website in-house, etc. It's not a bad thing, just set boundaries while you can before you're trying to rescue your infrastructure from whatever disaster befalls it.

# ? Sep 27, 2013 20:21

Precambrian Video Games: Aug 19, 2002

Thermopyle posted:

There's no magic software raid thing that works across Windows and Linux is there?

As a continuance to my last question about NTFS, I found that I have two 2TB drives, and as I'm not going to have anything not backed up on them, I was thinking about using RAID-0, but I'd also like to use them across Windows and Linux.

I'm guessing I'm going to have to use hardware RAID, but just thought I'd ask.

If you mean Intel RAID, you should be able to use it across both, although it's probably not going to be worth the headaches if one or the other drops and you don't actually need the performance gain. If you mean an actual controller, they're pretty expensive and again not worth it unless you actually really need it.

Getting ext working in Windows is infinitely more painful than NTFS in Linux. I recalling using some flaky drivers for ext2/3 years ago. I don't think ext4 drivers even exist yet.

# ? Sep 27, 2013 21:00

Adbot: ADBOT LOVES YOU

# ? Jun 12, 2024 22:03

spankmeister: Jun 15, 2008

NTFS on Linux is absolutely fine and has been for quite some time. Like years. You don't even need fuse anymore.

# ? Sep 27, 2013 21:06

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

«‹›1041 »