endolith's questions

X2Go with MATE desktop used to work fine, but after upgrading to 22.04 it has broken. If I try to connect with MATE (from Windows X2Go client), it shows a "New session started" popup, shows a blank window briefly, then closes. On the server, running dmesg --follow at the same time, I see:

[14238.559864] show_signal_msg: 3 callbacks suppressed
[14238.559867] mate-session[129320]: segfault at 7f3200000060 ip 00007f325d65f52f sp 00007fff9ee3b250 error 4 in libgdk-3.so.0.2404.29[7f325d623000+83000]
[14238.559876] Code: 57 82 fc ff 49 8b 9c 24 70 01 00 00 48 85 db 74 4e 49 89 c5 eb 0a 0f 1f 44 00 00 48 85 db 74 3f 48 8b 2b 48 89 de 48 8b 5b 08 <48> 8b 45 08 48 85 c0 74 e8 4c 39 e8 7f e3 49 8b bc 24 70 01 00 00

If I try to use something other than MATE, like "Connection to local desktop" (which also used to work), it shows "Select desktop" with :0 and :1 as options. If I select either, and hit "View Only" or "Full Access", I just get a black window and nothing in dmesg.

Same results from another Windows computer that used to work, so it's a server issue, not a client issue.

I have 3 physical external USB drives. All are btrfs. Two are combined into a single volume, the other is not. The two volumes are both in /etc/fstab to be mounted at /mnt and that works fine, but the combined one also keeps getting automatically mounted in /media, like a removable USB flash drive, and I don't know how to turn this off.

/etc/fstab:

# / was on /dev/sda6 during installation
UUID=619c1750-218e-4432-bb13-732a9c9192b5 /               ext4    errors=remount-ro 0       1
# /boot/efi was on /dev/sda1 during installation
UUID=3449-605A  /boot/efi       vfat    umask=0077      0       1
/swapfile                                 none            swap    sw              0       0

UUID=31abab38-ca1f-4541-9141-9506098d832a /mnt/butter auto nosuid,nodev,nofail,x-gvfs-show 0 0
UUID=3f8633ef-aad6-4444-87b6-831a11c3067e /mnt/Int_Butter auto nosuid,nodev,nofail,x-gvfs-show 0 0
UUID=2679ca57-e8e3-4b23-8574-69d751d62dbc /mnt/Risky auto nosuid,nodev,nofail,x-gvfs-show 0 0

All 3 are mounting to /mnt/ as they should, but butter is also being mounted to /media/ as if it were a USB thumb drive.

Often an automated process will completely fill up the main drive and I won't even be able to login over SSH to fix it. How do I prevent all the space from being used, so that I can always login to my headless server with SSH and delete something?

can't create socket: No space left on device

(There are an infinite number of things that could fill space. A good answer will not care which of them is doing it, but will prevent them from using the space that ssh needs to function.)

I despise these invisible scrollbars that don't exist until you hover over them and don't show you where you are on the page:

They are horrible to use over remote desktop.

How do I get real scrollbars? I'm using Mate desktop in Ubuntu 18.04

My SD card had a FAT32 partition and an EXT partition for Android. I don't need the EXT partition anymore, so I used GParted to delete it and increase the FAT32 to use the extra space, but Windows 7 doesn't see it anymore:

Ubuntu still sees it fine, though, including all the files:

Disks app "Repair filesystem" and GParted's "Check" did not fix it, either.

This is what parted shows:

(parted) print                                                            
Model: Generic- Multi-Card (scsi)
Disk /dev/sdb: 31.4GB
Sector size (logical/physical): 512B/512B
Partition Table: msdos
Disk Flags: 

Number  Start   End     Size    Type     File system  Flags
 1      4194kB  31.4GB  31.4GB  primary  fat32        lba

dosfsck says:

fsck.fat 4.1 (2017-01-24)
Checking we can access the last sector of the filesystem
Boot sector contents:
System ID "MSWIN4.1"
Media byte 0xf8 (hard disk)
       512 bytes per logical sector
     65536 bytes per cluster
       184 reserved sectors
First FAT starts at byte 94208 (sector 184)
         2 FATs, 32 bit entries
   1918976 bytes per FAT (= 3748 sectors)
Root directory start at cluster 6 (arbitrary size)
Data area starts at byte 3932160 (sector 7680)
    479620 data clusters (31432376320 bytes)
32 sectors/track, 64 heads
      8192 hidden sectors
  61399040 sectors total
Reclaiming unconnected clusters.
Checking free cluster summary.
/dev/sdb1: 92610 files, 408037/479620 clusters

Also tried TestDisk, it says:

TestDisk 7.1-WIP, Data Recovery Utility, April 2018
Christophe GRENIER <[email protected]>
https://www.cgsecurity.org

Disk /dev/sde - 31 GB / 29 GiB - CHS 3822 255 63
Current partition structure:
     Partition                  Start        End    Size in sectors

Invalid FAT boot sector
 1 P FAT32 LBA                0 130  3  3822 107 61   61399040
 1 P FAT32 LBA                0 130  3  3822 107 61   61399040
No partition is bootable

This drive doesn't need to be bootable. Quick Search finds the Linux partition I had deleted and resized over, but I guess that's not important:

TestDisk 7.1-WIP, Data Recovery Utility, April 2018
Christophe GRENIER <[email protected]>
https://www.cgsecurity.org

Disk /dev/sde - 31 GB / 29 GiB - CHS 3822 255 63
     Partition               Start        End    Size in sectors
>* Linux                 3692  99 56  3822 107 61    2088960 [Link2SD]

This seems more important:

TestDisk 7.1-WIP, Data Recovery Utility, April 2018
Christophe GRENIER <[email protected]>
https://www.cgsecurity.org

Disk /dev/sde - 31 GB / 29 GiB - CHS 3822 255 63
     Partition                  Start        End    Size in sectors
 1 P FAT32 LBA                0 130  3  3822 107 61   61399040

Boot sector
Bad

Backup boot sector
Bad

Second sectors (cluster information) are not identical.
Third sectors (second part of boot code) are not identical.

A valid FAT Boot sector must be present in order to access
any data; even if the partition is not bootable.

I tried Rebuild Boot Sector in TestDisk but it didn't work:

TestDisk 7.1-WIP, Data Recovery Utility, April 2018
Christophe GRENIER <[email protected]>
https://www.cgsecurity.org

Disk /dev/sde - 31 GB / 29 GiB - CHS 3822 255 63
     Partition                  Start        End    Size in sectors
 1 P FAT32 LBA                0 130  3  3822 107 61   61399040

FAT : 32
cluster_size 128 128
reserved     184 184
total_sect   61399040 61399040
fat32_length 3748 3748
root_cluster 2 6
free_count   uninitialised 71583
next_free    uninitialised 2
Extrapolated boot sector and current boot sector are different.

I set up a new computer with Ubuntu 16.04.1 LTS (GNU/Linux 4.4.0-45-generic x86_64)

It was working fine for a while but then started giving me address boundary errors and my Unity desktop theme screwed up, missing icons, etc.:

The only explanation I have is that I had enabled Nvidia and Intel proprietary drivers:

and it seems to have not happened since I turned them off. But the desktop is still screwed up and I don't trust anything else that may have been corrupted.

I ran destructive badblocks on all the HDDs before installing them and fsck seems fine on all of them. I ran memtest for a while and it didn't find any errors:

I can run sudo debsums -sa and it only finds files that I have intentionally changed. How do I confirm the cause and how do I repair any corrupted files?

I used Disks tool to make an image of a partition:

Then tried to mount it with Disk Image Mounter:

I enter my password twice:

Then it fails with this error message:

Unable to mount OLD SYSTEM_DRV

Error mounting /dev/loop1 at /media/endolith/OLD SYSTEM_DRV: Command-line `mount -t "ntfs" -o "uhelper=udisks2,nodev,nosuid,uid=1000,gid=1000" "/dev/loop1" "/media/endolith/OLD SYSTEM_DRV"' exited with non-zero exit status 21: ntfs-3g-mount: mount failed: Permission denied

Why is permission denied and how do I fix this?

In https://askubuntu.com/a/119818/5032 qbi says:

add noauto to your fstab:

...

So the filesystem will not automatically mounted at startup. Furthermore you write a script which basically mounts your mhddfs and add it at some late point in the boot process.

But how do you do this? My mergerfs never mounts all of its drives at startup, presumably because they aren't mounted before it? It's the last entry in fstab, like this (but with more drives):

/mnt/EXT4_drive/drivepool:/mnt/NTFS_drive/drivepool:/mnt/anotherdrive/drivepool /mnt/drivepool fuse.mergerfs defaults,allow_other,minfreespace=100M 0 0

When I run df after boot, the drive name only shows some of the mount points are included, and a fraction of the space it should have with all of them, so they aren't included if they aren't ready? I have to manually run sudo umount /mnt/drivepool ; sudo mount /mnt/drivepool after boot to fix it and show all of them.

Also computer has more than once booted into "Welcome to emergency mode!" when one of the drives was being automatically fscked while it tried to mount /mnt/drivepool. This hasn't happened lately, maybe because of a change in mergerfs that lets it ignore not-ready drives instead of halting because of them?

How would I automatically mount /mnt/drivepool ASAP after everything else in fstab has successfully mounted and fscked?

This has happened for as long as I can remember. Nautilus/Caja shows a shared icon on the folders, and they are accessible from Windows by typing in their folder name, but they don't show up in the list of folders.

Other folders do show up in the list, even though there's nothing different about the Sharing Options in Nautilus.

If I change the shared name, they show up. If I change it back, they disappear again.

Update:

I installed Ubuntu fresh on a new computer, shared two folders with caja-share with the same settings, and I can access one but not the other. Both have the same owner, group, and permissions:

drwxrwxrwx 2 endolith endolith 4.0K Sep 28 19:35 notwork/
...
drwxrwxrwx 2 endolith endolith 4.0K Sep 28 19:35 works/

the one that works is in /home, the other is on /mnt, so that's probably the reason?

It looks like the actual configuration for these are a bunch of text files under /var/lib/samba/usershares. /var/lib/samba/usershares/notwork is:

#VERSION 2
path=/mnt/drivepool/notwork
comment=
usershare_acl=S-1-1-0:F
guest_ok=y
sharename=notwork

/var/lib/samba/usershares/works is:

#VERSION 2
path=/home/endolith/works
comment=
usershare_acl=S-1-1-0:F
guest_ok=y
sharename=works

So the only difference is the path, as alluded to but not really explained in Here's my smb.conf file. Whats wrong with it?.

Mounting drivepool under my home folder doesn't fix it, either. So it's not the path itself but something to do with being inside the mount, which is mergerfs.

I am sharing the files using caja-share (mate-file-manager-share) under MATE desktop environment, Ubuntu 16.04.1 LTS, but it also has the standard Ubuntu environment installed.

Edit:

A clue! I found that Samba has log files for each machine that connects to it.

[2016/10/09 23:32:54.687515,  0] ../source3/param/loadparm.c:3259(process_usershare_file)
  process_usershare_file: stat of /var/lib/samba/usershares/win7backup failed. Permission denied
[2016/10/09 23:32:54.697705,  0] ../source3/smbd/service.c:199(set_current_service)
  chdir (/mnt/drivepool/Backups/win7backup) failed, reason: No such file or directory

There is such a directory, of course:

~> la /mnt/drivepool/Backups/
drwxrwxrwx 2 endolith endolith 4.0K Sep 26 22:42 win7backup/

Years ago, an Ubuntu update broke this and I've never found a solution.

• I have fish as my default shell.
• Ubuntu 12.10
• Locally (Gnome Terminal), Home and End keys work fine in both fish and bash.
• When I log in by SSH using PuTTY, then run bash, Home and End work fine inside of bash.
• However, when I log in by SSH using PuTTY, in the default fish shell, pressing Home key produces [1~ (sometimes erasing the line, sometimes not). When I press End, it produces [4~.
• If I use the default fish shell in Bitvise SSH client, Home and End work correctly.

How do I get Home and End to work correctly?

Is there a .deb file, repo, PPA, or anything to install CrashPlan other than the non-standard install.sh file? I have fish as my shell, and I had to mess around a lot and finally use chsh first before the install script would even work. I don't want to have to jump through all these hoops every time there's an upgrade.

I'm trying to set permissions for folders on USB drives that I've mounted with Nautilus, but when I change the permission, it just changes right back again. No error message or any other explanation. I've tried "Open as administrator" and it still doesn't work. How do I set the permissions so that they can be accessed as shared folders on the network?

Most of the time when I share a folder from Ubuntu, the name field is made yellow and it says "Share name too long". But even if I keep this name, it works fine. So what is it too long for? What is the limit?

Since no one can help me with sharing files from Samba, how can I reset it to the defaults and delete any residual configurations from previous releases? Then, presumably, the GUI sharing would work after I reinstall samba. I tried using dpkg reconfigure samba, but it made things even worse, and now my ubuntu computer's hostname stopped existing. >:(

In Ubuntu Maverick, I've shared some folders using the Nautilus "Sharing Options" GUI.

I can see them from Windows 7, but when I try to access them (from Windows) it asks for a username and password. No matter what I enter, it won't let me in. How do I configure this to share normally?

Update: I've found that some of the shared folders let me in, but others don't. Of the ones that do, some of their subfolders do, others don't, etc. How can I investigate what's causing this?

When I let the GUI install what it needs, it installs samba and libpam-smbpass. On the machine I am having trouble connecting to, libpam-smbpass is not installed. Could this be the problem?