Ask Your Question

zug234zwang's profile - activity

2019-02-27 15:01:07 -0500 commented answer Fedora 29 VirtualBox possible conflict

to sixpack13 : Did you misinterpret my query? The link that I indicated that I am about to follow will have me install

2019-02-27 09:32:30 -0500 asked a question Fedora 29 VirtualBox possible conflict

Fedora 29 VirtualBox possible conflict I did a complete re-install (not upgrade) to Fedora 29 in November 2018. At that

2019-01-04 01:21:05 -0500 commented question submitting feature request

Sorry for the ambiguity. I am reluctant to respond with a two page detailing of a very complex feature request. In sum

2019-01-04 00:24:36 -0500 received badge  Enthusiast
2019-01-03 21:02:44 -0500 asked a question submitting feature request

submitting feature request What is the procedure to follow to submit a feature request to the Fedora developers communit

2018-11-13 19:28:10 -0500 commented question could Fedora please reverse its policy re End-Of-Life

I internet-researched and discovered people having trouble upgrading.

2018-11-12 19:06:48 -0500 asked a question could Fedora please reverse its policy re End-Of-Life

could Fedora please reverse its policy re End-Of-Life Internet-researching, I've seen many situations where the user had

2017-12-21 12:34:44 -0500 received badge  Notable Question (source)
2017-09-01 03:11:59 -0500 received badge  Notable Question (source)
2017-09-01 03:11:59 -0500 received badge  Popular Question (source)
2017-05-06 07:45:37 -0500 received badge  Notable Question (source)
2017-05-06 07:45:37 -0500 received badge  Popular Question (source)
2016-05-26 12:02:45 -0500 commented question fedora 22 filesharing problem with virt-manager and Windows 7 guest

My bad for not being clear. Within the Windows 7 vm, if I add /mnt/disk1/ it does not show in the Windows network, so I can't map to it. Alternatively, if I add /myshareddisk1, which is (samba) "localhost-mounted", virtual machine manager (virt-manager) crashes.

2016-05-23 21:40:20 -0500 commented answer is there a way to install chrome on f22 after Aprill 2016?

netflix not work with firefox (Fedora 22 : cinnamon), but does work with chrome.

2016-05-23 19:49:47 -0500 asked a question fedora 22 filesharing problem with virt-manager and Windows 7 guest

I'm running Fedora 22 (64 bit, 16gb memory) with a cinnamon desktop and I would like to use Virtual Machine Manager to run Windows 7. However, I need filesharing between the Fedora 22 host and the virtual machine (Windows 7). I installed samba and Virtual Machine Manager via

sudo dnf install system-config-samba
sudo dnf install virt-manager

I then created the /mnt/disk1 (with 1 file) and /myshareddisk1 (empty) directories and used chmod 777 on both of them. Then I added the following trailer to /etc/samba/smb.conf:

comment = My Disk1
path = /mnt/disk1
public = yes
writeable = yes 

Then I used the following commands to enable the samba access to /mnt/disk1.

semanage fcontext -a -t samba_share_t "/mnt/disk1(/.*)?"
restorecon -R -v /mnt/disk1

Then, I used the following commands to create samba passwords for the root user, and my other user, steve.

smbpasswd -a root
smbpasswd -a steve

I then started samba and made /myshareddisk1 a mount of /mnt/disk1 via

systemctl start smb
mount //localhost/mydisk1 /myshareddisk1 -o user=steve

I then verified the samba process via both

smbclient -U steve -L localhost
ls /myshareddisk1 : the /mnt/disk1 file was displayed.

I then started the virtual machine service (systemctl start libvirtd) and started the virt-mgr application. Then I created a new virtual machine, and used a Windows 7 iso to install Windows 7 into this machine. The Windows 7 virtual machine runs ok.


Using the virt-mgr gui, I tried to add a samba file system to the Windows 7 virtual machine; an error message was generated.

Error starting domain: internal error: process exited while connecting to monitor: 
2016-05-23T23:54:21.548427Z qemu-system-x86_64: 
-device virtio-9p-pci,id=fs0,fsdev=fsdev-fs0,mount_tag=shared,bus=pci.0,addr=0x8: 
Virtio-9p Failed to initialize fs-driver with id:fsdev-fs0 and export path:/home/steve/shared02
2016-05-23T23:54:21.548491Z qemu-system-x86_64: 
-device virtio-9p-pci,id=fs0,fsdev=fsdev-fs0,mount_tag=shared,bus=pci.0,addr=0x8: 
Device 'virtio-9p-pci' could not be initialized

I then google researched, and discovered that (perhaps) virt-mgr doesn't like the source directory /home/steve/shared02, because it is not owned by the root user. I then tried a source directory of /mnt/disk1. This allowed virt-mgr to start the Windows 7 virtual machine, but did not provide any file sharing access. Then, I tried a source directory of /myshareddisk1, which (before the samba mounting) was owned by the root. This re-generated the error:

Error starting domain: internal error: process exited while connecting to monitor:

I then altered the mounting (mount //localhost/mydisk1 /myshareddisk1 -o user=root) and repeated the process: virt-mgr still produced the Error starting domain message. I then experimented within virt-mgr with various drive/mode combinations for the file system to be added - no joy.

Questions: If possible, please give direct answer(s) rather than referring to a webpage.

  1. I prefer a gui such as virtual ...

2016-04-20 12:08:33 -0500 received badge  Commentator
2016-04-20 12:08:33 -0500 commented answer fedora 22 cinnamon dmesg memory problem

thanks for the response, genodeftest. I could be mistaken, but I thought that my fedora22-kde4(dmesg) output did contain the "device node =" string. With that string missing, and a "...debugging..." message occurring, I questioned it. Also, on a cold boot, just after the bios message, when the screen shows Fedora options to select, the TOP ONE IN THE LIST is "fedora ... debug ..." which I also think is strange. This might relate to previous problems I had installing Fedy (i.e. repo not exist).

No problems were reported re sudo smartctl -s on -t short /dev/sdb.

2016-04-19 21:47:59 -0500 received badge  Nice Question (source)
2016-04-19 11:23:30 -0500 asked a question fedora 22 cinnamon dmesg memory problem

This query is probably more appropriately directed to a cinnamon or cinnamon-fedora forum, if any. However, I'll probably have better luck at this forum.

Boot disk recently died; decided to recover by re-installing everything, starting with Fedora 22 Live. kde proved problematic so switched to cinnamon. Several months ago, while using Fedora 22 with kde4, I researched how to make an external hard disk bootable. One of the steps was to find the device node assigned to the drive.

dmesg | tail
[ 4783.165535] sd 6:0:0:0: [sdc] Attached SCSI disk
device node = sdc

Today, I'm under fedora 22 64 bit with cinnamon 2.6.13, 16gb memory. I inserted a hard disk to my usb connected docking station (device sdd). Cinnamon automatically mounted it; I then unmounted the drive, leaving it still assigned to device sdd. Then:

[  474.183592] scsi 7:0:0:0: Direct-Access     WDC WD30 EZRX-00DC0B0          PQ: 0 ANSI: 5
[  474.185430] sd 7:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
[  474.185608] sd 7:0:0:0: [sdd] 5860533168 512-byte logical blocks: (3.00 TB/2.73 TiB)
[  474.185861] sd 7:0:0:0: [sdd] Write Protect is off
[  474.185864] sd 7:0:0:0: [sdd] Mode Sense: 28 00 00 00
[  474.186112] sd 7:0:0:0: [sdd] No Caching mode page found
[  474.186115] sd 7:0:0:0: [sdd] Assuming drive cache: write through
[  474.186293] sd 7:0:0:0: Attached scsi generic sg4 type 0
[  474.188831] sd 7:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
[  474.274863]  sdd: sdd1 sdd2
[  474.275939] sd 7:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
[  474.276534] sd 7:0:0:0: [sdd] Attached SCSI disk
[  840.116816] DMA-API: debugging out of memory - disabling
[  920.589012] kworker/dying (16) used greatest stack depth: 10848 bytes left
[ 1093.678988] kworker/dying (6) used greatest stack depth: 10672 bytes left
[ 1223.786129] usb 3-10: USB disconnect, device number 5

The heading lines from the top command (executed as the super user) are:

top - 08:38:49 up 26 min,  2 users,  load average: 0.00, 0.07, 0.17
Tasks: 224 total,   1 running, 223 sleeping,   0 stopped,   0 zombie
%Cpu(s):  0.8 us,  0.5 sy,  0.0 ni, 98.4 id,  0.2 wa,  0.0 hi,  0.0 si,  0.0 st
KiB Mem : 16271316 total, 14042704 free,  1125872 used,  1102740 buff/cache
KiB Swap:  8191996 total,  8191996 free,        0 used. 14745112 avail Mem 


  1. (In linux-newbie terms) what does this message mean: "DMA-API: debugging out of memory - disabling"?

  2. Why is there no "device node = sdd" message in the dmesg output? Does it relate to the "debugging ... memory..." message?

  3. With 16gb memory, I installed from Fedora 22 Live without making any effort to tweak my memory settings. Do the memory lines from the top command indicate a ...

2016-04-14 21:22:06 -0500 commented answer fedora 22 dnf upgrade failed to sync

The debate seems to be "more that can go wrong" vs "ease of use". Since I am a programmer, and dnf is straightforward (and powerful), it is a no-brainer for me to avoid Fedy unless I'm forced into it; [e.g. if I can't find a workable Adobe Flash repo any other way, I'll query Linux forum(s)]. For true Linux newbie, ease of use" might be attractive.

2016-04-14 21:20:59 -0500 commented answer fedora 22 dnf upgrade failed to sync

thanks florian, very interesting response. following up, i googled : fedora "not use fedy", and fedora "against fedy"; interesting reading. I was merely following the advice in both and .

2016-04-14 21:00:49 -0500 received badge  Scholar (source)
2016-04-14 12:55:56 -0500 asked a question fedora 22 dnf upgrade failed to sync


sudo dnf update : excerpt

[steve@ip98-167-78-229 lucky_you.2007]$ sudo dnf upgrade
Failed to synchronize cache for repo 'fedy' from 
Cannot download repomd.xml: Cannot download repodata/repomd.xml: 
All mirrors were tried, disabling.
Last metadata expiration check performed 0:47:49 ago on Thu Apr 14 07:33:10 2016.
Dependencies resolved.
 Package                  Arch      Version       Repository    Size
 dracut                   x86_64    041-15.fc22   updates       324 k
 dracut-config-rescue     x86_64    041-15.fc22   updates        45 k

Transaction Summary
Upgrade  21 Packages

Total download size: 7.9 M
Is this ok [y/N]: 


fedora 22 64 bit, cinnamon 2.6.13, 16gb memory


Boot disk recently died; decided to recover by re-installing everything, starting with Fedora 22 Live. kde proved problematic so switched to cinnamon. Then had trouble installing fedy. With the (partial or full) failures shown below, I (mistakenly) did not retain the error messages.

The following seemed to totally fail:

su -c "curl -o fedy-installer && chmod +x fedy-installer && ./fedy-installer"

The following seemed to either partially or totally fail. I noticed what seemed to be error messages.

went to
selected the (green) download button
downloaded fedy-installer
in a terminal : 
  made file executable (chmod 777)
  then executed it : ./fedy-installer

Because of the error messages in the previous approach, I then:

google searched : fedora 22 install fedy
found and played a youtube video :
the video pointed to
on that webpage the following command was given:  
bash -c 'su -c "curl -o fedy-installer && chmod +x fedy-installer && ./fedy-installer"'

The above seemed to succeed, "Fedy" was added to the cinnamon start Menu. In fact, (partially unrelated) after a previous (terminal) attempt to install Adobe Flash failed, the Fedy installer successfully installed it. This was verified via a webpage that requires Flash.

Long Term

Actually, I've been "living off of" a usb-connected external hard disk, using it as a boot disk. New hard disks coming in a few days; then I'll (again) re-install from scratch, starting with Fedora 22 Live.


  1. Why is sudo dnf upgrade yielding an error message? (In linux-newbie terms) what does the message mean? Was the message (plausibly) caused by my 2nd Fedy installer attempt, executing ./fedy-installer?

  2. With fedy apparently successfully installed and dnf upgrade apparently working correctly, is the error message "harmless"?

  3. Assuming no re-install, is there a way of removing the error message (prefer a terminal command, rather than low-level editing of some file on the root partition).

  4. Assuming that I re-install from scratch and bypass the two failed attempts to install fedy, should I expect the error message to not appear?

2016-03-09 21:01:59 -0500 received badge  Popular Question (source)
2016-03-03 16:53:26 -0500 asked a question recovery by package re-install

In case of hard disk failure, I want my recovery to (ideally) re-create the pre-hard-disk-crash environment. After researching backup strategies for my Fedora 22 home pc, I've decided to only use rsync on my /home and /etc directories, and to (during recovery) manually re-install packages and manually upgrade packages. After examining the dnf documentation and recovery-related forum entries, I'm confused about the explicit recovery steps to take. The following chart shows the number of entries for various commands:

00042_lines dnf history list
00026_lines dnf history userinstalled
49112_lines dnf list all
02542_lines dnf list installed
02505_lines rpm -qa

With no linux recovery experience, my tentative strategy is shown below: I REQUEST CRITICISM FROM EXPERIENCED LINUX USERS. I ALSO REQUEST SPECIFIC OPTIONS TO USE WITH THE "dnf [options] upgrade" COMMAND (STEP # 5 BELOW).

(1) In addition to using rsync to maintain an independent copy of my /home and /etc directories, maintain a copy of dnf_history_userinstalled.txt, et al. Ignore the other 4 entries above.

(2) Maintain a bootable flashdrive of Fedora 22 Workstation Live, updating it whenever updates the corresponding iso.

To recover when the (boot) hard disk fails:

(3) Use the flashdrive to install the "full" Fedora 22 to a new hard disk.

(4) Use dnf_history_userinstalled.txt to manually install the (less than 50) packages.

(5) Execute the "dnf [options] upgrade" command. Hopefully, since I kept the flashdrive of Fedora 22 Workstation Live "current", this process will be "streamlined".

(6) Use rsync to update the /home and /etc directories on the new hard disk.

2016-02-12 21:12:24 -0500 commented answer rsync on root vs package reinstall

I don't think the previous answer tackled any of the three specific issues that I raised. Rsync vs clone is a separate issue; I'm happy enough with rsync. Also, I feel that these issues are best served by explicit responses, rather than referencing another webpage, unless that webpage specifically discusses/debates one of the three issues that I am raising.

2016-02-11 01:10:09 -0500 received badge  Supporter (source)
2016-02-11 00:52:10 -0500 asked a question rsync on root vs package reinstall

This is a followup of what I believe to be three open issues re the following query on this forum:

rsync boot disk dirs to exclude

Because of my inexperience, one of my replies on that query became mis-positioned, which caused that webpage to be confusing. I'm hoping for responses from several people on this followup query. Issues:

(1) Suppose that:

a. My boot drive is (device = sda) and my docking station houses a backup hard drive (device = sdc).

b. I have already used a recent version of Fedora22 Live to install the (full) Fedora22 on (device = sdc), making sdc bootable.

c. I want to maintain sdc as a full backup (i.e. a functional "clone") in case sda goes bad, totally relying on periodic rsync's of the sda-root against sdc to keep sdc current.

d. I want to be able to both boot the backup from the docker (position sdc) or replace the original boot disk with the backup, booting the backup in (position sda).

e. I want to TOTALLY AVOID having to reinstall any packages.

My Fedora 22 boot drive has the following directory structure:

.autorelabel, bin, boot, .config, dev, etc, home, lib, lib32, lib64, lost+found, media, mnt, opt, proc, root, run, sbin, srv, sys, tmp, usr, var

My tentative list of directories to exclude is:

/dev, /home, /media, /mnt, /proc, /run/media, /sys, /tmp, /var/lock, /var/run

/home is excluded because I rsync it separately. Under the assumptions of issue (1), I REQUEST DISCUSSION OF DIRECTORIES TO EXCLUDE.

(2) In my Linux-inexperience, and focusing on supposition (1)d above, I have speculated that /etc/fstab and /boot/grub2/grub.cfg may need special handling. My intent is:

a. On sda, copy /etc/fstab to .../fstab_sda and copy /boot/grub2/grub.cfg to .../grub_sda.cfg.

b. Exclude /etc/fstab and /boot/grub2/grub.cfg from the rsync, HOPING that the backup will therefore boot successfully from (position sdc).

c. If I want to replace the boot disk with the backup (booting the backup from position sda), I will replace .../fstab and .../grub.cfg with their ..._sda counterparts, HOPING that the backup will therefore boot successfully from (position sda).


(3) My research SUGGESTS that (1) + (2) above are overkill, and that I should abandon the rsync_on_the_root strategy in favor of:

a. Only rsync'g /home and /etc.

b. Maintaining a list of installed packages with dnf.

c. Restoring by using Fedora 22 Live to place the (full) Fedora 22 on a new disk, manually rsync'g /home and /etc on this new disk, and then using dnf to manually re-install the packages to this new disk.

d. Periodically checking for a newer version of the Fedora 22 (workstation) Live iso.

On my system, "dnf history userinstalled" reports only 24 packages since my ... (more)

2016-02-10 23:31:39 -0500 commented answer rsync boot disk dirs to exclude

Thanks again for the responses, Florian. My bad re generating a confusing web page. Anyway, the only pending issues are having other people suggest other Fedora 22 dirs to exclude, and the (partly generic, partly Fedora 22 specific) issue of package_re-install vs rsync against the root. I will consolidate the remaining issues into a separate query; LET'S CONSIDER THIS QUERY (WEBPAGE) CLOSED.

2016-02-10 08:47:44 -0500 edited answer rsync boot disk dirs to exclude

The web page "re-located" my "answer", which was my response to Florian's comments. Logically, my answer should immediately precede the comment: "Here is how I would do it. Your plan sounds good..."

Below given as an answer because I didn't want to split my reply into multiple comments.

Very interesting comments, thanks florian. Goals:

a. A full backup of my harddrive in case my main one goes bad.

b. Ability to boot the backup hard drive from either the docking station (device = sdc) or as a replacement hard disk (device = sda). This gives me flexibility in a restore situation and facilitates "rolling back" the main boot drive (using device=sdc to restore device=sda), in case a "bad" package is installed, without having to open my pc case.

c. After booting with the backup harddrive, all installed packages should be present; no need to re-install packages.

After reading your comments I have two followup questions:

(1) re goal b above, and noting that /boot will be included and /dev excluded from the rsync, I believe that I will need special handling for /etc/fstab and /boot/grub2/grub.cfg. I will probably create fstab_original and grub_original.cfg on my main boot disk (just before each rsync) and then specifically exclude /etc/fstab and /boot/grub2/grub.cfg from the rsync. This should allow the backup to boot from the docking station (device = sdc). If I ever need to swap the backup for the original boot (i.e. move the backup into device=sda), I will first boot Fedora 22 live from a separate flash drive and replace fstab and grub.cfg with their ..._original counterparts, on the backup.

Is this workable/practical? Are there other areas that will also need special handling?

(2) re goal c above, I was surprised to encounter the following recommendation:

rpm -qa | tee installed_packages.list

I presumed that all installed packages would automatically be replicated by the rsync. Doesn't that make the above command obselete? What am I missing?

2016-02-09 17:44:11 -0500 commented answer rsync boot disk dirs to exclude

The web page "re-located" my "answer", which was my response to Florian's comments. Logically, my answer should immediately precede the comment: "Here is how I would do it. Your plan sounds good..."

2016-02-09 17:20:38 -0500 commented answer rsync boot disk dirs to exclude

(4) Continuting the last question, assume that the last package installed is bad, I have uninstalled via (dnf history undo ...), the resulting system is non-bootable, and I successfully repair the system. Does this mean that the repaired system approximates the system just before the last package was installed? If not, is the difference "significant"?