My university uses Cisco anyconnect and Duo to authenticate. When logging into the vpn, I have to run anyconnect, connect to the vpn server, and type in:
userid password and a second factor 6-digit code from Duo.
The whole process is slow and annoying. Is there anything I can do to turn this into a command line operation? Can I use any open source vpn, is there a way to record my password and to get authentication at least to just be clicking yes on the duo app?
I tried to copy partitions from a 500GB internal M.2 SSD to an external USB3.x enclosure around an M.2 2TB drive.
My first attempt used an A to C cable, and the command:
dd if=/dev/nvme01p3 of=/dev/sda1 bs=1M
took approximately 6.5 hours for a 240GB partition. It was reporting 10MB/s. This seemed really slow. It turned out that the port I was connecting on was probably USB2, so I used the short USB C to C cable, and got 40MB/s, and it took about 1.5 hours for a similar partition.
dd should be pretty low level, particularly if the files in question are partitions. Isn't that just raw block on the disk? So when in another question, some answers suggested I use copy partition in gparted. There were mistakes anyway, so I tried that, and was stunned when it achieved what I thought it should, which was more than 1GB/sec. 4 minutes later, the partition was redone.
The question is, how is it possible that dd can be 40x slower than gparted? If this question has to move to stackoverflow, I will do it, but it's in a gray area between software and linux.
cgdb is a better version of gdb using curses to draw colored text and show the current executed line. gdb works, but every time the program prints out text, gdb gets shifted, and you have to press control-L to refresh.
I have many students using cgdb under windows, but for some odd reason I cannot run it myself on 4 separate ubuntu machines. So it seems like the default configuration for Ubuntu 20.04LTS just doesn't work. Here is a picture of what happens before the terminal locks up:
CGDB: a curses debugger
version 0.6.7
type q<Enter> to exit
type help<Enter> for GDB help
type <ESC>:help<Enter> for CGDB help
*qqqqqqqj[39m(B(B[m[8;40H(0[0;1;7m[36m[39m(B(B[m[16;80H[H[2J(0[0;1;7m[36ml
qqqqqqqqqqqk[39m(B(B[m[2;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m[3
6mx[39m(B(B[m[3;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m[36mx
[39m(B(B[m[4;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m[36mx[39m(B(B
[m[5;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m[36mx[39m(B(B[m[6;1H(0[0;1;7m
[36mx[39m(B(B[m[80G(0[0;1;7m[36mx[39m(B(B[m[7;1H(0[0
;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m[36mx[39m(B(B[m[8;1H(0[0;1;7m
[36mx[39m(B(B[m[8;17H(0[0;1;7m[36m[39m(B(B[m[ No Source Available ]
[80G(0[0;1;7m[36mx[39m(B(B[m[9;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0[0;1;7m
[36mx[39m(B(B[m[10;1H(0[0;1;7m[36mx[39m(B(B[m[80G(0
--Type <RET> for more, q to quit, c to continue without paging--
So this works in Windows using the msys2 shell, and a student running arch linux has no problem either. Thinking that somehow the default terminal is the problem, I installed alacritty, and it happens there too.
TERM is set to xterm-256color, could that be the problem? If so, what do I set it to?
My desktop is running ubuntu 20.04LTS It recently mysteriously started to popup a little alert window at the top informing me that a dialog is ready instead of just popping up the dialog. If I don't click in time, the dialog is somewhere in back. I suspect it's some look and feel parameter that accidentally got set, allowing requesters to pop up in the back so they are not seen, and it's highly annoying.
I know it happens in vscode. I believe it happens in many programs, so I think it's system-wide, not just the editor.
Suspend used to work flawlessly. Then it stopped. I thought the machine was dead, but was able to ssh in. Eventually I realized that if I just unplugged the monitor, and plugged it back in the screen would go live again.
Oddly, sometimes it works anyway. No idea why. The only other symptom is that my system went from 60 Hz to 29.something. I was just in settings and saw that, so I put it back.
nvidia-smi shows 460.80. I suppose I could update to 460.84.
Is there any way I can diagnose the problem?
I have a machine running Ubuntu 16.04. It has a 2TB drive with a lot of files, but certainly more than 30% free. Everything was working. I was running qbittorrent. I shut the machine down and didn't bring it up again for a few weeks. After rebooting, qbittorrent fails after about 60 seconds saying there are too many open files.
cat /proc/sys/fs/file-max
shows 713000
sysctl fs.file-nr
fs.file-nr = 9760 0 781241
sudo lsof | wc -l
lsof: WARNING: can't stat() fuse.gvfsd-fuse file system
/run/user /1000/gvfs
Output information may be incomplete.
133915
Under qbittorrent is 17,116 directories, 167826 files. So there are quite a lot. The program crashes on startup so I cannot query it for how many files are open.
This seems like a horrible bug in qbittorrent. Even if they are loading files, they should not open 171k all at once.
Whatever is wrong, I have the same number of files I did when I shut the system down.
This problem is right at the edge of askubuntu and stackoverflow, so if you think it should move, I would appreciate help in moving it.
On ubuntu, Netbeans 10 is not working. On windows, the same version of netbeans works. On both platforms, the code will not compile with jdk11, but at least can use jdk8 on windows.
I installed netbeans under ubuntu 19:
sudo apt install netbeans
Here is what netbeans about shows:
Product Version: Apache NetBeans IDE 10.0 (Build 20190125-debian-10.0)
Java: 11.0.4; OpenJDK 64-Bit Server VM 11.0.4+11-post-Ubuntu-1ubuntu219.04
Runtime: OpenJDK Runtime Environment 11.0.4+11-post-Ubuntu-1ubuntu219.04
System: Linux version 5.0.0-23-generic running on amd64; UTF-8; en_US (nb)
User directory: /home/user/.netbeans/10.0
Cache directory: /home/user/.cache/netbeans/10.0
Netbeans will not build using jdk11. I downloaded jdk-8u221-linux-x64.tar.gz
and installed to /opt/jdk1.8.0_221
then if the project library is to set jdk1.8:
the code compiles, but still does not run. Under the run tab, it says it is using "project default" and when I click, it is always using the original system jdk11. I can select jdk1.8, but when I close that requester it never uses it.
To verify, I compile on the command line.
javac *.java
AssignmentList.java:1: error: package com.sun.javafx.collections does
not exist
import com.sun.javafx.collections.ListListenerHelper;
An empty program with a blank main does work, jdk11 merely does not work with swing, not sure what other libraries.
If I compile manually using jdk1.8, it works.
/opt/jdk1.8.0_221/bin/javac *.java
/opt/jdk1.8.0_221/bin/java WhateverProgram
How do I get netbeans to compile java under Ubuntu, and what is wrong with JDK11?
Ransomware attacks could use zero-day exploits, but often an attacker will just fool a gullible user into running an executable by downloading and clicking.
Suppose we have a naive user and want to restrict them to the normal path. Is there any way to restrict them from creating a file with executable privilege?
Or, more generally, is there any way to build an access control list and define that this user may only execute files in this list?
On Ubuntu 19.04, Libreoffice seems to be there, but when I clicked, all the apps seem to be missing (like the word processor and the spreadsheet). So I first installed OpenOffice. Then I figured out that I could install Libreoffice as well:
sudo apt install libreoffice
so I now can verify that both have the same issue.
Both Openoffice and Libreoffice have an option claiming to digitally sign a document.
So I created a certificate using instructions from this site: https://websiteforstudents.com/self-signed-certificates-ubuntu-17-04-17-10/ Here is the shell script I wrote to do it so I would not forget the steps:
#! /bin/bash
# see https://websiteforstudents.com/self-signed-certificates-ubuntu-17-04-17-10/
name=dbk
openssl genrsa -aes128 -out $name.key 2048 #generate a key
#best to generate without a passphrase, so next command removes it
openssl rsa -in $name.key -out $name.key
# create a certificate signing request. This one is good for a year
days=365
openssl req -new -days $days -key $name.key -out $name.csr
openssl x509 -in $name.csr -out $name.crt -req -signkey $name.key -days $days
#make it readable only to you, to protect it
chmod 400 $name.*
OpenOffice still didn't find the certificate or offer me any way to select a directory. In the wiki for openoffice, I found a page claiming there are 4 ways to select the directory where OpenOffice will look: https://wiki.openoffice.org/wiki/How_to_use_digital_Signatures
One of those ways would be to set an environment variable:
export MOZILLA_CERTIFICATE_FOLDER=~/cert
This didn't work either. In Libreoffice, I can see an option to start a certificate manager. When I click, I get the following window.
I can click on the certificate in the cert directory but the open button is not clickable.
How can I get either word processor to find the certificate that I created in ~/cert, or is there something wrong with the steps I used to create it?
I am running Ubuntu 18.10 on an i7-8 with 32Gb RAM and an NVIDIA RX-2080. We installed and x failed after a while. I had installed NVIDIA drivers. It seemed like the nouveau driver started running, no idea why. A couple of weeks later the resolution suddenly went to hell one morning, and the machine is barely responding.
Now it happened again. We blacklisted the nouveau driver, still didn't work, re-installed NVIDIA. Now that the machine works again, can anyone explain why this happens, and whether blacklisting is likely to solve this problem? Is there anything else I can do?
While trying to install my usual environment on Ubuntu 18.10 I have run into many, many problems.
I am curious why packages that are supposedly working under 18 seem to have all these issues?
I managed to build emacs from source, so my problems with emacs-gtk/emacs-common/emacs26 can all be put on the back burner.
But clementine and simplescreenrecorder both install using apt install and both, when run complain about
error while loading shared libraries: libdouble-conversion.so.1: cannot open shared object file: No such file or directory
Reading package lists... Done
Building dependency tree
Reading state information... Done
libdouble-conversion-dev is already the newest version (2.0.1-5).
0 upgraded, 0 newly installed, 0 to remove and 9 not upgraded.
From what I can see, there is a symbolic link:
lrwxrwxrwx 1 root root 25 Apr 10 2018 /usr/lib/x86_64-linux-gnu/libdouble-conversion.so -> libdouble-conversion.so.1
to a nonexistent file libdouble-conversion.so.1
apt-cache policy libdouble-conversion1
libdouble-conversion1:
Installed: 2.0.1-5
Candidate: 2.0.1-5
Version table:
*** 2.0.1-5 500
500 http://us.archive.ubuntu.com/ubuntu cosmic/main amd64 Packages
100 /var/lib/dpkg/status
I also built libdouble-conversion manually, but it creates /usr/local/lib/libdouble-conversion.so.3 and clementine, simplescreenrecorder, and octave are all expecting .so.1
I deleted libdouble-conversion is /usr/local/lib and tried:
sudo apt install --reinstall libdouble-conversion1
now octave, simplescreenrecorder and clementine all coredump. Is there any way to identify what is happening? To show which shared library is failing? I would like to show that it is in fact, still libdouble-conversion.
I would of course like to solve this problem, but I would also appreciate if anyone could explain why there are so many problems like this with 18.10?
Installing emacs under 18.10 (Cosmic Cuttlefish) fails
When I search for libotf, it is there:
locate libotf.so
/usr/lib/x86_64-linux-gnu/libotf.so
So why does fixing the install fail?
sudo apt install -f
Reading package lists... Done
Building dependency tree
Reading state information... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
2 not fully installed or removed.
After this operation, 0 B of additional disk space will be used.
Setting up emacs-gtk (1:25.2+1-11) ...
Install emacsen-common for emacs
emacsen-common: Handling install of emacsen flavor emacs
emacs: error while loading shared libraries: libotf.so.0: cannot open shared object file: No such file or directory
ERROR: install script from emacsen-common package failed
dpkg: error processing package emacs-gtk (--configure):
installed emacs-gtk package post-installation script subprocess returned error exit status 1
No apport report written because the error message indicates its a followup error from a previous failure.
dpkg: dependency problems prevent configuration of emacs25:
emacs25 depends on emacs-gtk (>= 1:25); however:
Package emacs-gtk is not configured yet.
dpkg: error processing package emacs25 (--configure):
dependency problems - leaving unconfigured
Errors were encountered while processing:
emacs-gtk
emacs25
E: Sub-process /usr/bin/dpkg returned an error code (1)
Things rapidly spiralling from bad to worse on this disasterous Dell Inspiron 15 5577.
First, Windows got corrupted, twice, so I decided to install Ubuntu. Didn't have a USB handy and it doesn't work on SD cards, so I went to buy a thumb drive quickly. I got a 64Gb drive which is bigger than it wants, possibly a problem? I had installed Windows media on the thing, and it created a 32 Gb partition. To do this I wiped out everythign and used Startup Disk Creator to write the partition onto the drive.
First tried to UEFI and secure boot on, since the ubuntu utlity has been creating UEFI drives for some time (I thought). BIOS did let me switch boot order, and the drive showed up with Ubuntu 18. That didn't work, so I re-installed with Ubuntu 16.04 LTS. That did not work either. It crashes during boot, the logo freezes in the splash screen.
I switched off secure boot. Then based on the following discussion I switched off UEFI, even though I thought the drive is UEFI.
https://ubuntuforums.org/showthread.php?t=2382295
The booting and BIOS are getting slower all the time, and I am going to try to reset to factory defaults. I don't understand how the BIOS can get slow -- it should not be doing anything. It seems to be waiting for non-existent drives.
The other posting mentioned "fastboot" but I cannot find that option anywhere.
The first time it booted off the UEFI USB drive but crashed. Now, I can't even get BIOS to find the USB drive so I can try to boot off it.
I would love to get to the point again where it fires up the grub menu, but now I'm not even getting to that point.
Any idea what is going wrong?
rhythmbox never added my music. I thought it might be something wrong with the mp3 plugin, but vlc and the default movie player can play the mp3s. I uninstalled rhythmbox and reinstalled in case it was a discovery issue on startup.
So in desperation, I installed deepin. That does not work either. It now appears that neither one can see my music directory.
I have / mounted on an SSD and /big mounted on a large drive neither program can see the drive.
Sure enough, when I copy to the main drive, rhythmbox can see the files.
There are no special permissions on these files. They are owned by me. How does rhythmbox not work with /big/Music ?
We have a network of 12 machines and one server at the moment connected by a gigabit switch in a single room. We want every user to be able to log in and use their account on any machine, as well as being able to execute mpi jobs where each machine appears identical. This implies mounting a common home drive and managing accounts.
I know NFS from way back. I just tested one that I suspect is insecure.
/etc/exports
Following the directions for Ubuntu's NFS configuration does things the old way. But I saw in the exports file gss/krb5i which leads me to believe there is a secure, encrypted way to do this.
We also found freeipa which seems to be a complete packaged solution. But it's difficult to know what the other alternatives may be. Can anyone point to an overview listing alternatives, or state from experience a good secure way to achieve common user ids and files across a network?
Thanks!
On a server, I have mounted a big drive on /home2 in order to store all the user accounts.
Supposedly in order to set the default directory to /home2, we are supposed to edit /etc/default/useradd
Inside is a line (commented out)
#HOME=/home
which I change to:
HOME=/home2
Yet subsequently, useradd -m continues to create users home directories in /home. For the moment, I just mounted the new drive as /home, and that did work obviously.
Can anyone tell me what I am doing wrong?
I have a Dell Optiplex 9020 which was running Ubuntu 14.04LTS. Ubuntu complained that it was out of date and no longer providing security updates. Of course, it did not offer a way to upgrade, but looking it up on stackexchange, I found the following advice:
Can I smoothly upgrade from one LTS to next LTS release?
sudo apt-get do-release-upgrade
Now the machine locks up after booting. I can't see anything in a log. There is a text cursor. Alt-F* does not work to get to a shell.
I can press the power button, and it then shows the ubuntu logo before shutting down in a controlled manner. So Linux is running.
Any ideas as to how to recover my system? I am assuming I should type some options in grub.
Booting with Linux 3.16.0-77-generic (recovery mode)
Recovery mode shows:
Started user manage for UID 112
started automatic USB/bluetooth printer setup...
Failed to start Network manager wait online
At this point ctrl-alt-del does not work, but pressing the power button does continue through the sequence and shut down.
I recently tried to install a few packages on ubuntu 14.04 and they failed. At first I thought it was just one package, but it is a general failure of apt-get. For example:
$ mono
The program 'mono' is currently not installed. You can install it by typing:
sudo apt-get install mono-runtime
and when I try to install:
$ sudo apt-get install mono-runtime
Reading package lists... Done
Building dependency tree
Reading state information... Done
You might want to run 'apt-get -f install' to correct these:
The following packages have unmet dependencies:
linux-signed-image-3.16.0-62-generic : Depends: linux-image-3.16.0-62-generic (= 3.16.0-62.83~14.04.1) but 3.16.0-62.82~14.04.1 is to be installed
mono-runtime : Depends: mono-runtime-sgen (= 3.2.8+dfsg-4ubuntu1.1) but it is not going to be installed
E: Unmet dependencies. Try 'apt-get -f install' with no packages (or specify a solution).
I thought perhaps 14.04 is in end of life so I looked into upgrading. But the updater crashes with a requester:
Check if you are using third party repositories. If so disable them, since they are a common source of problems.
Furthermore run the following command in a Terminal: apt-get install -f
I unclicked and removed all third party repositories. Same error. How can I uncorrupt? What is even wrong?
In Fedora 21, we can type:
apt-get ins
hit tab, and it will complete the word install
We are using an instance of Ubuntu 10.04 on a cloud platform, and tab completion works for files, but does not work for command parameters.
Is there any way to get this to happen on Ubuntu 10.04? We have to stick for compatibility with software running in this virtual machine.
NVIDIA complains that it cannot install with x running.
All I want to do is kill the X server, but I can't do that. If I kill it, it simply restarts. I found an article claiming I could change the grub parameter of the kernel to:
text_$vt_handoff
but this doesn't work with 13.04, it boots as usual into X.
I tried init 1, but it warned me not to install in single user mode either.
I would probably prefer the proprietary NVIDIA driver, though this process is difficult and seems much, much worse than it used to be under Fedora Core 4, a giant step backward in ease of installation. Why is it so hard to come by instructions for something so mainstream? Why is it now so hard to shut down the X server?
As an alternative I would be willing to use the nvidia_current package, but want to be able to use dual screens, twinview to nvidia. Does the open software driver support that? Does it also support rotating monitors by 90 degrees?