Tweaking Fedora
by Beartooth
Within the next three weeks or so, barring any great surprises, I
hope to be putting Fedora 27 onto four PCs, a laptop, and a netbook. IIRC,
the netbook, which I haven't used recently, has F24 now; all the others
are running 26. I'll do all I can by upgrading, because every time I do a
fresh install, it costs me something more than a whole day per machine to
do the tweaking I need.
(I've been running Fedora since it was RedHat7; so my trifocal
fingers and arthritic eyeballs know their jobs IF I get all the tweaks
right -- a great boon and a bringer of cyber-survival, since I also keep
getting slower and more forgetful....)
A correspondent on another list says one used to be able to use
Mondo Rescue to grab all the settings on an existing install and clone
them onto a new one. That would save me vast tedium.
But the Mondo Rescue site lists only rpms for Fedora 23 and
before. And either I'm garbling my correspondent's directions, or they
don't work any more -- or both.
I tried a few variations on "dnf install Mondo-xyzq". I also
downloaded a few .rpms from Mondo's repository and ran "rpm -ivh" against
them. Both tries failed.
Is there a tutorial somewhere? Has Mondo Rescue forked into
something with another name? Have the Fedora Gurux and Alpha Plus
Technoids come up with a replacement while I wasn't looking??
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
--
/home/btth/sig/nqc
1 year, 8 months
Resolve fc27.i686 vs. fc27.x86_64??
by Beartooth
Probably in connection with some version of yumex, I have somehow
gotten both */fc26.i686 and */fc26.x86_64 forms of a dozen or so rpms for
compiz.
This has not been a problem till now. fc26 (and, unless I'm
misremembering, fc25 and fc24 before that) has just ignored the
superfluous ones. But now system-upgrade is choking on the ones for fc27.
The machine is actually .x86_64, as I have now discovered. I
tried adding --allowerasing and --allowerasing --best, but with no
success.
Do I dare do something like "dnf remove compiz"?? (Or perhaps
"dnf remove */compiz/*/.i686 ???)
1 year, 8 months
Re: NetworkManager-wait-online is still utterly, and completely, broken
by James Hogarth
On 19 Dec 2017 07:49, <Francis.Montagnac(a)inria.fr> wrote:
On Mon, 18 Dec 2017 14:12:55 -0800 Gordon Messmer wrote:
> On 12/17/2017 01:17 PM, Francis.Montagnac(a)inria.fr wrote:
>> Dec 17 21:32:59 X systemd[1]: Mounting /data2...
>> Dec 17 21:33:19 X mount[996]: mount to NFS server 'Y' failed:
Resource temporarily unavailable, retrying
>> Dec 17 21:33:19 X systemd[1]: Mounted /data2.
> Can you reproduce that condition and then get the output of "systemctl
> status NetworkManager-wait-online"? That should confirm that the
> ExecStart process does not include the -s flag.
Here it is:
systemctl status NetworkManager-wait-online
. NetworkManager-wait-online.service - Network Manager Wait Online
Loaded: loaded
(/usr/lib/systemd/system/NetworkManager-wait-online.service;
disabled; vendor preset: enabled)
Active: active (exited) since Tue 2017-12-19 07:58:05 CET; 4min 12s
ago
Docs: man:nm-online(1)
Process: 743 ExecStart=/usr/bin/nm-online -q --timeout=30
(code=exited, status=0/SUCCESS)
Main PID: 743 (code=exited, status=0/SUCCESS)
Tasks: 0 (limit: 4915)
CGroup: /system.slice/NetworkManager-wait-online.service
Dec 19 07:57:58 X systemd[1]: Starting Network Manager Wait Online...
Dec 19 07:58:05 X systemd[1]: Started Network Manager Wait Online.
but still:
journalctl -b -u data2.mount
-- Logs begin at Thu 2015-04-09 18:41:55 CEST, end at Tue 2017-12-19
08:01:55 CET. --
Dec 19 07:58:05 X systemd[1]: Mounting /data2...
Dec 19 07:58:25 X mount[1036]: mount to NFS server 'Y' failed: Resource
temporarily unavailable, retrying
Dec 19 07:58:25 X systemd[1]: Mounted /data2.
The setup of this server is:
eth0 static IP declaration with dns servers and gateway
eth0.2006 static IP declaration: additionnal vlan interface
with the associated port on a cisco switch configured as:
switchport trunk native vlan 390
switchport trunk allowed vlan 390,2006
switchport mode trunk
switchport nonegotiate
storm-control broadcast level 5.00 3.00
no cdp enable
/etc/resolv.conf was generated by NetworkManager at the same second
indicated by nm-online:
stat /etc/resolv.conf
File: /etc/resolv.conf
Size: 76 Blocks: 8 IO Block: 4096 regular file
Device: 806h/2054d Inode: 19924685 Links: 1
Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2017-12-19 07:58:05.539740447 +0100
Modify: 2017-12-19 07:58:05.437735660 +0100
Change: 2017-12-19 07:58:05.438735707 +0100
Birth: -
The 'Y' server is in the eth0 subnet.
'Y' has to be resolved by the DNS.
As said before, even pinging the gateway IP succeeds only more than
20s after the routes (as seen by "ip route") are defined.
Weird no?
Cisco switch and that delay for traffic on a static address?
Sounds like symptoms of STP delaying s port from listening to forwarding...
Have you looked at disabling on that interface or setting portfast if it's
only host facing?
1 year, 8 months
Samba support fails in F27
by Temlakos
Everyone:
After several days of fruitless effort I have to bring this to the
community. Since I upgraded to F27, I have never been able to sustain
any viable network connection to other computers on my network that run
Windows 10. Which means if I ever have to transfer files between or
among them, I have to use USB flash drives or even USB-connectable
moveable HDD's.
Now I suppose I could buy a spare Western Digital Passport for this
application. But I would like to know what I'm missing.
For the record:
The computer I'm typing this on, that runs F27, presently has a wireless
connection to the network. I hope, within the next two months, to move
to an environment where I can connect this computer to the network using
a MoCA adapter, a switch in the room, and a coaxial connector in the
wall to connect this computer, a printer, and maybe a laptop if I bring
it into the same room. That never used to make any difference, but yes,
I'm going to try to "remedy" that "fault" (if anyone here chooses to
find fault with using a wireless connection).
The Samba workgroup is named Home. That's case-sensitive.
I have User authentication on the Samba server.
I created Samba user accounts for every Fedora user account on this
machine--all three of them.
With F26, I could always look up Home under Samba Shares and find it.
But now--nothing. And even specifying Home gets a not-found-here kind of
message.
The Services application shows that both smb and nmb are running.
So what's wrong, where might the fault lie, and how do I correct it?
Temlakos
1 year, 8 months
Questions regarding dual booting with Windows 8.1
by Sudhir Khanger
Hello,
I have to install Windows 8.1 so that Dell technician can troubleshoot my system. I have not installed Windows in years, so, I wanted to confirm the process before destroying my system.
I will create a 50GB partition for Windows 8.1. I installed Fedora with automatic partitioning. As a result, home partition which I will resize is in middle. Is it okay to resize the partition in the middle to install Windows?
$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme0n1 259:0 0 238.5G 0 disk
├─nvme0n1p1 259:1 0 200M 0 part /boot/efi
├─nvme0n1p2 259:2 0 1G 0 part /boot
├─nvme0n1p3 259:3 0 197.3G 0 part /home
├─nvme0n1p4 259:4 0 20G 0 part [SWAP]
└─nvme0n1p5 259:5 0 20G 0 part /
Once I have installed Windows, it will probably overwrite Grub and possibly UEFI entries. Do I simply have to reinstall the Grub for Fedora to start detecting it as a second OS.
grub2-install /dev/nvme0n1
grub2-mkconfig -o /boot/efi/EFI/fedora/grub.cfg
Do I have the process correctly understood?
1 year, 8 months
piping
by Patrick Dupre
Hello,
It is not a fedora question, but I am sure that somebody can help me.
I do:
my_sort files.asc > new_file.asc
and then
gawk -f file.awk -v sampl=T new_file.asc
to get what I want.
how can I do the same in a single command with a pipe (and not generating
the new_file.asc) ?
Thank.
===========================================================================
Patrick DUPRÉ | | email: pdupre(a)gmx.com
Laboratoire de Physico-Chimie de l'Atmosphère | |
Université du Littoral-Côte d'Opale | |
Tel. (33)-(0)3 28 23 76 12 | | Fax: 03 28 65 82 44
189A, avenue Maurice Schumann | | 59140 Dunkerque, France
===========================================================================
1 year, 8 months
Re: NetworkManager-wait-online is still utterly, and completely,
broken
by Francis.Montagnac＠inria.fr
broken
In-reply-to: Your message of "Sun, 17 Dec 2017 08:20:37 -0500."
<20171217082037.2148ddd8@zooty>
On Sun, 17 Dec 2017 08:20:37 -0500 Tom Horsley wrote:
> On Sun, 17 Dec 2017 11:07:52 +0000
> Dr J Austin wrote:
>> Anyone else with the same problem/
> I gave up on getting NFS mounts to work a long time ago.
I had the same problem when (finally) deciding to reconfigure a server
to use NetworkManager instead of the old network service.
I manage to make it work with a "drop-in" configuration file for
NetworkManager-wait-online.service that replaces the call of nm-online
by waiting for all the wanted routes to be up and that a gateway is
reachable. See attachment.
For this server it's really long, and waiting for the routes is not
sufficient at all:
Dec 17 12:46:37 xxx bash[1047]: Network Manager Wait Online routes
took 9 seconds
Dec 17 12:47:07 xxx bash[1047]: Network Manager Wait Online gateway
took 30 seconds
Dec 17 12:47:07 xxx systemd[1]: Started Network Manager Wait Online.
Francis
1 year, 8 months
config files
by Beartooth
Is there a list, or a way to get a list, of Fedora's config
files? Better yet, how about just those files I've tweaked??
1 year, 8 months
firefox-57.0.1-1.fc27.x86_64 -
by Bob Goodwin
Upon booting this morning I found that I had been updated to
firefox-57.0.1-1.fc27.x86_64. I have been keeping Firefox downgraded [to
version 52 whatever] with dnf upgrade --exclude firefox. Dnf now refuses
to downgrade below version 57.
Am I really stuck with this or is there a way around it?
Bob
1 year, 8 months
