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 »

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

evol262 posted:

DG/UX 5.4 on a system which still had reel-to-reel. SunOS 4.1.4. IRIX. The first company I worked for had a lot of legacy hardware running legacy meteorology stuff they hadn't ported yet. The SunOS box was owned by Thompson Reuters (but fed from our datasources, and under our management).

In about the same way as VMware ran RHEL. They do, but it's not at all stock, and their filesystem stuff is totally custom.

IBM likewise sells SONAS as GPFS in a box, and while it's all RHEL on the surface, the guts of the product are their cluster management and filesystem layers. Makes it really easy to make support mad when you start adding cronjobs and other poo poo they explicitly tell you not to do but that the product is useless without.

# ? Oct 21, 2013 16:06

Adbot: ADBOT LOVES YOU

# ? Jun 12, 2024 23:49

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

eightysixed posted:

I'm not sure where to ask this question, so I'll start here - If there's a better place, let me know and I'll try there. With that being said, I'm having a strange intranet problem. We have a Linux box (running Debian 6) in the office that acts as the company's Samba share server as well as running a project management software that's accessible via browser. It locally runs Apache2 serving activeCollab to all users on the network. It is a small network; Only serving six (6) users. The issue is this -

Out of nowhere, the server becomes completely inaccessible from any other workstation on the network. No Samba shares, mounted drives, HTTP, or even SSH access - But when I ping the IP of the server, it responds, albeit a bit slow, which is even more confusing. However when this happens/no accessibility, the server itself has complete internet connectivity, which blows my mind. Any idea why this is this all of a sudden happening? - if network connectivity was lost, I wouldn't be able to look our from the server itself. Nothing has changed whatsoever. So when I try to SSH, or access the webserver, or even retrieve files via a samba share, the server shows completely dead. But siting at the server itself has complete network connectivity. What could it be?

I've switched ethernet cables, tried different ports, etc, etc. If this

Pinging 192.168.1.100 with 32 bytes of data:
Reply from 192.168.1.100: bytes=32 time=4ms TTL=64
Reply from 192.168.1.100: bytes=32 time=570ms TTL=64
Reply from 192.168.1.100: bytes=32 time=184ms TTL=64
Reply from 192.168.1.100: bytes=32 time=110ms TTL=64

tcpdump. See whether or not you get an answer to telnet on the public IP from another box. Then from the server. Then check localhost from the server. You need to troubleshoot the network before we can begin to help. Figure out where in the chain it's failing.

# ? Oct 21, 2013 16:57

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

SurgicalOntologist posted:

Does anyone have any recomendations for CLI music players? All the ones I found so far are actually terminal interfaces which isn't what I'm looking for. I'd like to be able to do something like play ~/path/to/album or play -r --shuffle ~/path/to/artist ~/path/to/anotherartist. I'm more interested in options like 'recursive' and 'shuffle' than any fancy audiophile filters and stuff.

VLC comes with cvlc and/or rvlc It's what I use.

code:

cvlc --random filespec

for instance. Most of those will also work with vlc or any of the other vlc variants like nvlc (ncurses) or whichever the minimal gui version is, which i forget.

# ? Oct 21, 2013 19:15

xtal: Jan 9, 2011; by Fluffdaddy

mpd & mpc is pretty wonderful

# ? Oct 21, 2013 19:41

eightysixed: Sep 23, 2004; I always tell the truth. Even when I lie.

evol262 posted:

tcpdump. See whether or not you get an answer to telnet on the public IP from another box. Then from the server. Then check localhost from the server. You need to troubleshoot the network before we can begin to help. Figure out where in the chain it's failing.

tcdump looks completely normal to me :confused:

I can see all attempted inbound connections. IPTables isn't even configured, so that's not it. I've killed fail2ban, but nothing. This problem just started occurring out of nowhere.

That's the strangest thing about the whole thing. telnetting to both 80 and 22 work fine (as well as pinging the local IP), even when everything else doesn't. The server itself is completely network connected, it will just all of a sudden become inaccessible from anywhere other than local. Apache isn't throwing any errors, but really, thats a minor point. When it becomes inaccessible, I can't SSH into the box at all, even though I can go sit at the terminal and have COMPLETE connectivity. It just seems intermittent all of a sudden for no reason. No, no other servers on the network, and no switches or routers have been plugged in on top of themselves. Literally, nothing has changed - I can't figure it out.

I have to be missing something exceptionally dumb. :smith:

# ? Oct 21, 2013 20:04

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

i've had something like this happen before.

Do you have a nic with an identical MAC address on the network?

# ? Oct 21, 2013 20:23

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

eightysixed posted:

tcdump looks completely normal to me I can see all attempted inbound connections.

Ok, but is Apache responding? Or sshd? If you start either in the foreground in debug mode, what happens when you try to connect?

eightysixed posted:

That's the strangest thing about the whole thing. telnetting to both 80 and 22 work fine (as well as pinging the local IP), even when everything else doesn't. The server itself is completely network connected, it will just all of a sudden become inaccessible from anywhere other than local. Apache isn't throwing any errors, but really, thats a minor point. When it becomes inaccessible, I can't SSH into the box at all, even though I can go sit at the terminal and have COMPLETE connectivity. It just seems intermittent all of a sudden for no reason. No, no other servers on the network, and no switches or routers have been plugged in on top of themselves. Literally, nothing has changed - I can't figure it out.

So, what's the actual resolution when this happens? Reboot? "Socket is open but no data crosses" sounds like something is hung up on the server. Debug mode.

# ? Oct 21, 2013 20:56

spankmeister: Jun 15, 2008

Try pinging with increasing packet sizes (and DF bit set) this could be mtu related.

# ? Oct 21, 2013 20:58

robostac: Sep 23, 2009

It sounds like it could be a duplicate ip or mac issue - while pinging from something that can't connect, try unplugging the server and see if pings continue?

# ? Oct 21, 2013 21:37

eightysixed: Sep 23, 2004; I always tell the truth. Even when I lie.

Alright, so it's currently happening right now. It seems to be getting worse and worse :confused:

For reference for the rest of this post, a client terminal is 192.168.1.123. The server is 192.168.1.100

evol262 posted:

Ok, but is Apache responding? Or sshd? If you start either in the foreground in debug mode, what happens when you try to connect?

Trying to connect via SSH while it's going tits up yields the following:

code:

kevin@somedomain:/var$ sudo /etc/init.d/ssh stop Stopping OpenBSD Secure Shell server: sshd.
kevin@somedomain:/var$ sudo /etc/init.d/ssh start -d Starting OpenBSD Secure Shell server: sshddebug1: sshd version
OpenSSH_5.5p1 Debian-6+squeeze3
debug1: read PEM private key done: type RSA
debug1: Checking blacklist file /usr/share/ssh/blacklist.RSA-2048
debug1: Checking blacklist file /etc/ssh/blacklist.RSA-2048
debug1: private host key: #0 type 1 RSA
debug1: read PEM private key done: type DSA
debug1: Checking blacklist file /usr/share/ssh/blacklist.DSA-1024
debug1: Checking blacklist file /etc/ssh/blacklist.DSA-1024
debug1: private host key: #1 type 2 DSA
debug1: rexec_argv[0]='/usr/sbin/sshd'
debug1: rexec_argv[1]='-d'
Set /proc/self/oom_adj from 0 to -17
debug1: Bind to port 22 on 0.0.0.0.
Server listening on 0.0.0.0 port 22.
debug1: Bind to port 22 on ::.
Server listening on :: port 22.
debug1: Server will not fork when running in debugging mode.
debug1: rexec start in 5 out 5 newsock 5 pipe -1 sock 8
debug1: inetd sockets after dupping: 3, 3 Connection from 192.168.1.123 port 5694

evol262 posted:

So, what's the actual resolution when this happens? Reboot? "Socket is open but no data crosses" sounds like something is hung up on the server. Debug mode.

There really isn't one. Restarting the whole server doesn't really affect anything. When it happens and I try to connect via SSH over putty/kitty I get: Network error: Software caused connection abort. A tcpdump listening for that IP yields the following while this is happening:

code:

kevin@somedomain:~$ sudo tcpdump -n -i eth2 tcp port 22 and host 192.168.1.123
[sudo] password for kevin:
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on eth2, link-type EN10MB (Ethernet), capture size 65535 bytes
16:22:11.609826 IP 192.168.1.123.60181 > 192.168.1.100.22: Flags [P.], seq 1182202841:1182202909, ack 174904417, win 16212, length 68
16:22:11.612588 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:12.040346 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:12.896344 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:14.608346 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:18.032338 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:24.880340 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:38.576355 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:23:05.968352 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68

# ? Oct 21, 2013 22:22

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

eightysixed posted:

There really isn't one. Restarting the whole server doesn't really affect anything. When it happens and I try to connect via SSH over putty/kitty I get: Network error: Software caused connection abort. A tcpdump listening for that IP yields the following while this is happening:

So you re-image it? You buy a new one from your VAR every time it happens? You power cycle the whole building? It goes away randomly?

quote:

code:

kevin@somedomain:~$ sudo tcpdump -n -i eth2 tcp port 22 and host 192.168.1.123
[sudo] password for kevin:
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on eth2, link-type EN10MB (Ethernet), capture size 65535 bytes
16:22:11.609826 IP 192.168.1.123.60181 > 192.168.1.100.22: Flags [P.], seq 1182202841:1182202909, ack 174904417, win 16212, length 68
16:22:11.612588 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:12.040346 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:12.896344 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:14.608346 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:18.032338 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:24.880340 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:22:38.576355 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68
16:23:05.968352 IP 192.168.1.100.22 > 192.168.1.123.60181: Flags [P.], seq 1:69, ack 68, win 159, length 68

So you can reach the server. The server just can't reach you back. You're on the same subnet. Is the subnet mask wrong? Does it respond to broadcast pings? Do broadcast pings see the rest of the network? Do you have a managed switch you can port monitor or port mirror on?

# ? Oct 21, 2013 22:32

jre: Sep 2, 2011; To the cloud ?

eightysixed posted:

Alright, so it's currently happening right now. It seems to be getting worse and worse

robostac posted:

It sounds like it could be a duplicate ip or mac issue - while pinging from something that can't connect, try unplugging the server and see if pings continue?

Have you tried what robostac suggested ? pull the network cable from the server and see if you still get a response to pings? The most likely reason for this suddenly occurring is something else on your network being assigned the same ip or having a duplicate mac. Have you checked that the server entry listed by "arp -a" on the client matches the mac address for the network card in the server.

# ? Oct 21, 2013 22:56

eightysixed: Sep 23, 2004; I always tell the truth. Even when I lie.

robostac posted:

It sounds like it could be a duplicate ip or mac issue - while pinging from something that can't connect, try unplugging the server and see if pings continue?

jre posted:

Have you tried what robostac suggested ? pull the network cable from the server and see if you still get a response to pings? The most likely reason for this suddenly occurring is something else on your network being assigned the same ip or having a duplicate mac. Have you checked that the server entry listed by "arp -a" on the client matches the mac address for the network card in the server.

It was the new receptionist's cell phone. The server is set to static 192.168.1.100. All other workstations also have a static IP. Unplugged the server and was still was able to ping. She turned off her phone, no ping. DHCP was giving her phone the .100 for whatever reason. I can't believe I didn't even think about that. Thank you both. I knew it was something so simple that I just wasnt getting it. I feel ashamed :ughh:

evol, enjoy your shindig :cheers:

# ? Oct 21, 2013 23:34

Pockyless: Jun 6, 2004; With flaming Canadians and such :(

Is it possible to rename a SCSI generic device file in Debian(ie, /dev/sg1 -> /dev/sg0)? Sometimes when I replace a bad drive, the sg driver adds the new device file to /dev/sg2 even if sg0 doesn't exist anymore. This makes it kind of annoying when I want to do things that rely on the old generic device name, like automating smartctl checks.

# ? Oct 22, 2013 00:44

3spades: Mar 20, 2003; 37! My girlfriend sucked 37 dicks!

Customer: In a row?

Are you using udev? Per debian's hotplug guide section 9.11.4, you may be able to make a static rule for it if udev is what is handing out the assignment of the device. Assuming you can make a rule by its iqn or something (check with ls -l /dev/disk/by-path/ ).

# ? Oct 22, 2013 16:00

Superdawg: Jan 28, 2009

Does anyone have a way of telling if and why xen is holding an lvm volume open?

I've got this device-mapper volume that shows an open count of 1, despite nothing on the host OS indicating that anything is holding it open.

code:

Name                       Maj Min Stat Open Targ Event  UUID
xenstore-lfarm--004--extra 252  30 L--w    1    1      0 LVM-5iBtw4CHqCieEn5HtcrMmhPaWefyJEjbDvrXvdqb6qr0gmzkxpXaHKrRGv5nG1s9

# ? Oct 23, 2013 15:12

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

In Chef, how come SSL certificates get stored in plaintext in the certificates folder? Don't you have to be protective of the crt and key files?

# ? Oct 23, 2013 19:59

polar: Nov 3, 2003

Ive been interested in using linux. I have a macbook that has 2 video cards and ubuntu isnt able to install drivers to use my card/cards properly. it crashes after i install the recomended drivers on linux for a nvidea card. tried both sets of drivers and both crash.

I was wondering if there was a release of linux that would be able to handle running or switching video cards?

# ? Oct 23, 2013 19:59

spankmeister: Jun 15, 2008

fletcher posted:

In Chef, how come SSL certificates get stored in plaintext in the certificates folder? Don't you have to be protective of the crt and key files?

Well sure but at some point your going to have to read them. You can encrypt them but then the program using the certs needs to know the key to decrypt so you'll have to store that in plain text somewhere.

# ? Oct 23, 2013 20:02

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

spankmeister posted:

Well sure but at some point your going to have to read them. You can encrypt them but then the program using the certs needs to know the key to decrypt so you'll have to store that in plain text somewhere.

Yea it's gotta be stored in plain text on the server, but does it also need to be stored in plain text in my git repo?

# ? Oct 23, 2013 20:05

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

fletcher posted:

Yea it's gotta be stored in plain text on the server, but does it also need to be stored in plain text in my git repo?

Why are you storing the certs in your Git repo? You can just recreate them if you ever lose them.

# ? Oct 23, 2013 20:11

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

polar posted:

Ive been interested in using linux. I have a macbook that has 2 video cards and ubuntu isnt able to install drivers to use my card/cards properly. it crashes after i install the recomended drivers on linux for a nvidea card. tried both sets of drivers and both crash.

I was wondering if there was a release of linux that would be able to handle running or switching video cards?

Try this. Generally, though, nVidia's told Linux users they won't ever write switchable drivers for Linux, so Bumblebee is the best you get.

# ? Oct 23, 2013 20:18

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

Misogynist posted:

Why are you storing the certs in your Git repo? You can just recreate them if you ever lose them.

I guess since I'm using chef-solo and I'm trying to use cookbook_file to deploy the crt & key files for my HTTP server to use, so I have to add them to the git repo. That way when chef-solo runs on the server it can use cookbook_file to deploy the key & crt files.

How does recreating them work? I generated the key & csr, send it off to the SSL vendor for signing, and they give me back a crt. I thought if I regenerated the key & csr, it wouldn't be compatible with the crt they provided me?

# ? Oct 23, 2013 20:26

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

fletcher posted:

I guess since I'm using chef-solo and I'm trying to use cookbook_file to deploy the crt & key files for my HTTP server to use, so I have to add them to the git repo. That way when chef-solo runs on the server it can use cookbook_file to deploy the key & crt files.

How does recreating them work? I generated the key & csr, send it off to the SSL vendor for signing, and they give me back a crt. I thought if I regenerated the key & csr, it wouldn't be compatible with the crt they provided me?

I had actually assumed you were talking about the client certs used to communicate with a chef-server. But if you really want, you can store your cert data in an encrypted data bag, or use something like chef-vault.

# ? Oct 23, 2013 20:40

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

evol262 posted:

Try this. Generally, though, nVidia's told Linux users they won't ever write switchable drivers for Linux, so Bumblebee is the best you get.

They basically don't want to stray from the FOSS stack too too much, and the FOSS stack doesn't support dual-chip right now. We're working on it with the new RandR version and the new DRM stuff we're working on (render nodes, etc.), and they're fine with supporting it as long as we do most of the gruntwork.

# ? Oct 23, 2013 21:08

Baron Bifford: May 24, 2006; Can't post for 2 years!

When I installed Windows 8 on my PC, it created a 350 MB "system reserved" partition. What is this for? Can I safely wipe it? I'm asking this because I want to install Ubuntu on the same drive and am thinking of using that partition as the boot partition.

# ? Oct 24, 2013 19:16

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

Baron Bifford posted:

When I installed Windows 8 on my PC, it created a 350 MB "system reserved" partition. What is this for? Can I safely wipe it? I'm asking this because I want to install Ubuntu on the same drive and am thinking of using that partition as the boot partition.

Any particular reason you need to install Ubuntu on bare metal? It's so much easier to just spin up a Virtual Machine from inside Windows using something like VirtualBox.

# ? Oct 24, 2013 19:31

Baron Bifford: May 24, 2006; Can't post for 2 years!

This way it works faster. Also, that doesn't answer my question.

# ? Oct 24, 2013 19:44

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

Got another chef question:

Right now I'm using the mysql cookbook and just relying on whatever version of MySQL is in the apt repo for Ubuntu 12.04. As an easy way of installing a different version, I was thinking of just uploading a .deb to an s3 bucket, and then grab it using remote_file before I include_recipe "mysql::server". I'll also have specified the name of that .deb in node["mysql"]["server"]["packages"]. Is this a dumb idea? Is there a better way I should be doing this?

# ? Oct 24, 2013 19:48

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

Baron Bifford posted:

This way it works faster. Also, that doesn't answer my question.

VMs are really really fast if your CPU supports virtualization technologies. I can't perceive any performance difference between my VM and my host OS. I don't know the answer to your question, but it sounds Windows specific and you're in the Linux thread.

Unless you're trying to run games on Linux, then don't use a VM

# ? Oct 24, 2013 19:53

spankmeister: Jun 15, 2008

Baron Bifford posted:

When I installed Windows 8 on my PC, it created a 350 MB "system reserved" partition. What is this for? Can I safely wipe it? I'm asking this because I want to install Ubuntu on the same drive and am thinking of using that partition as the boot partition.

You cannot delete it as it contains the boot partition with the boot loader and all the files and stuff. If you were to use bitlocker that's also where the unencrypted part lives that decrypts the actual partition when booting.

But in short: don't delete it lest you end up with an unbootable windows.

# ? Oct 24, 2013 19:56

SamDabbers: May 26, 2003

Baron Bifford posted:

When I installed Windows 8 on my PC, it created a 350 MB "system reserved" partition. What is this for? Can I safely wipe it? I'm asking this because I want to install Ubuntu on the same drive and am thinking of using that partition as the boot partition.

Windows Vista/7/8 uses that small partition as a boot partition. Don't format or overwrite it. IIRC it's FAT32 so you might be able to stick a kernel/initrd there in addition.

efb

# ? Oct 24, 2013 19:56

Baron Bifford: May 24, 2006; Can't post for 2 years!

Ubuntu will need to install a new boot manager, since the Win8 boot manager doesn't support Linux. Could I wipe that boot partition and place Ubuntu's bootloader there?

# ? Oct 24, 2013 20:14

SamDabbers: May 26, 2003

Baron Bifford posted:

Ubuntu will need to install a new boot manager, since the Win8 boot manager doesn't support Linux. Could I wipe that boot partition and place Ubuntu's bootloader there?

I'd just make a separate /boot for Ubuntu, install GRUB on that, and configure GRUB to chainload the Windows bootloader from its existing partition. Just set the Ubuntu boot partition to active so GRUB is loaded first.

Here's an article about dual-booting Linux and Windows 8 with UEFI/GPT. It's written for Arch, but the advice should apply for Ubuntu as well.

SamDabbers fucked around with this message at 20:19 on Oct 24, 2013

# ? Oct 24, 2013 20:15

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

fletcher posted:

Got another chef question:

Right now I'm using the mysql cookbook and just relying on whatever version of MySQL is in the apt repo for Ubuntu 12.04. As an easy way of installing a different version, I was thinking of just uploading a .deb to an s3 bucket, and then grab it using remote_file before I include_recipe "mysql::server". I'll also have specified the name of that .deb in node["mysql"]["server"]["packages"]. Is this a dumb idea? Is there a better way I should be doing this?

I think most people interested in pinning specific package versions should be running their own apt repos and pulling down whatever packages they need. It gives you more control (you don't need to worry about old versions going missing), your local mirror is going to be faster, and it cuts down on the number of external dependencies that might be broken when you go to install a new system. Most of the apt repo management software out there is a trainwreck built for managing repositories with tens of thousands of packages. You'll be fine with something that reads your files and dumps a Packages index.

I use a really simple tool called prm to manage our (small) local repositories, and then I sync them to S3 with s3cmd. Where it's necessary, I pin specific versions using the apt_preference LWRP in the apt cookbook.

# ? Oct 24, 2013 20:16

spankmeister: Jun 15, 2008

Baron Bifford posted:

Ubuntu will need to install a new boot manager, since the Win8 boot manager doesn't support Linux. Could I wipe that boot partition and place Ubuntu's bootloader there?

Yes but then you will only be able to boot ubuntu.

You really need to keep that partition, or reinstall windows without the partition.
(With 7 you could trick the installer not to create that partition, not sure if you can with 8)
Besides, what is 350mb anyway these days?

# ? Oct 24, 2013 20:20

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

fletcher posted:

VMs are really really fast if your CPU supports virtualization technologies. I can't perceive any performance difference between my VM and my host OS. I don't know the answer to your question, but it sounds Windows specific and you're in the Linux thread.

Unless you're trying to run games on Linux, then don't use a VM

If you think GUI performance under VirtualBox is the same as bare metal, when VBox multi-monitor support even works right at all, your eyes clearly run at 10 Hz and you should see a neurologist.

# ? Oct 24, 2013 20:21

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

Misogynist posted:

when VBox multi-monitor support even works right at all

This is still better than VMware Fusion's multi-monitor "support"

# ? Oct 24, 2013 20:35

SurgicalOntologist: Jun 17, 2004

SamDabbers posted:

I'd just make a separate /boot for Ubuntu, install GRUB on that, and configure GRUB to chainload the Windows bootloader from its existing partition. Just set the Ubuntu boot partition to active so GRUB is loaded first.

Here's an article about dual-booting Linux and Windows 8 with UEFI/GPT. It's written for Arch, but the advice should apply for Ubuntu as well.

Alternatively, you can install GRUB2 into your Ubuntu / partition, then install EasyBCD in Windows 8 and tell it to add your Linux install to the W8 bootloader (it's quite easy, it will find it automatically). Then you can keep the pretty Windows 8 bootloader and still dual-boot.

The downside is it's a fake bootloader - that is, when you choose Ubuntu it will reboot into GRUB2, so if boot times are important to you this isn't a great solution. On the upside, the W8 bootloader is nicer looking than GRUB.

# ? Oct 24, 2013 20:37

Adbot: ADBOT LOVES YOU

# ? Jun 12, 2024 23:49

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

evol262 posted:

This is still better than VMware Fusion's multi-monitor "support"

Have you used 4.3.0? Attempting to maximize any window moves it to your leftmost monitor.

# ? Oct 24, 2013 20:43

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

«‹›1041 »