VPN options
by Keith Lofstrom
I am planning on running a Virtual Private Network from my Fedora
firewall out to a UML virtual colo (running RH9) at another site.
That site will be the place I present services to the world;
httpd, ssh, sftp, smtp. This is to comply with the "no servers"
and dynamic ip restrictions on my Comcast connection to the net;
if my firewall always drives an outbound connection to the
colocation site, I am not worried about changes of ip address,
and I am not opening any inbound ports.
There are a number of options for the VPN - the most attractive
are cipe ( http://sites.inka.de/sites/bigred/devel/cipe.html )
and FreeSwan ( http://www.freeswan.org/ ), though I am told that
one can do all this through an ssh tunnel. I would rather have
simple and secure than super-duper; I have plenty of bandwidth,
and will send outbound http and smtp from the firewall, so the
main bandwidth user will be incoming spam/b/b/b/b mail.
Anyone have some experiences to share about setting up VPN? Is
there anything about either cipe or FreeSwan that is likely to
break with FC1 or FC2?
Keith
--
Keith Lofstrom keithl(a)ieee.org Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs
1 year, 2 months
ACPI errors (was: f24 boot fails; need help).
by William
Good evening,
One of the problems reported in the log file "rdsosreport.txt" generated
by the failed boots discussed in the "f24 boot fails; need help" topic
was a set of "ACPI" errors. Here are the relevant lines (with added
line numbers) from that log file (the first and last lines are just for
context):
----------
862 [ 1.218683] coyote kernel: ata4: SATA link up 1.5 Gbps
(SStatus 113 SControl 300)
863 [ 1.218927] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
864 [ 1.219196] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT1._GTF] (Node ffff92ef8e0db488), AE_NOT_FOUND
(20160930/psparse-543)
865 [ 1.219856] coyote kernel: ata2.00: ATA-8:
ST2000DM001-1CH164, CC24, max UDMA/133
866 [ 1.220017] coyote kernel: ata2.00: 3907029168 sectors, multi
16: LBA48 NCQ (depth 31/32), AA
867 [ 1.220812] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
868 [ 1.221100] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT1._GTF] (Node ffff92ef8e0db488), AE_NOT_FOUND
(20160930/psparse-543)
869 [ 1.221620] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
870 [ 1.221909] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT5._GTF] (Node ffff92ef8e0dbac8), AE_NOT_FOUND
(20160930/psparse-543)
871 [ 1.222200] coyote kernel: ata6.00: ATAPI: HL-DT-ST BD-RE
BH14NS40, 1.00, max UDMA/100
872 [ 1.222374] coyote kernel: ata2.00: configured for UDMA/133
873 [ 1.222582] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
874 [ 1.222908] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT3._GTF] (Node ffff92ef8e0db348), AE_NOT_FOUND
(20160930/psparse-543)
875 [ 1.222919] coyote kernel: scsi 1:0:0:0: Direct-Access
ATA ST2000DM001-1CH1 CC24 PQ: 0 ANSI: 5
876 [ 1.223466] coyote kernel: ata4.00: ATAPI: PIONEER BD-RW
BDR-208M, 1.10, max UDMA/100
877 [ 1.226360] coyote kernel: ata7: SATA link down (SStatus 0
SControl 300)
878 [ 1.226407] coyote kernel: ata9: SATA link down (SStatus 0
SControl 300)
879 [ 1.226501] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
880 [ 1.226507] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT5._GTF] (Node ffff92ef8e0dbac8), AE_NOT_FOUND
(20160930/psparse-543)
881 [ 1.226519] coyote kernel: ata6.00: configured for UDMA/100
882 [ 1.227197] coyote kernel: ACPI Error: [DSSP] Namespace
lookup failure, AE_NOT_FOUND (20160930/psargs-359)
883 [ 1.227204] coyote kernel: ACPI Error: Method parse/execution
failed [\_SB.PCI0.SAT0.SPT3._GTF] (Node ffff92ef8e0db348), AE_NOT_FOUND
(20160930/psparse-543)
884 [ 1.227217] coyote kernel: ata4.00: configured for UDMA/100
----------
What is this all about? Is this a real problem or a false alarm? How
do I fix it?
thanks,
Bill.
2 years
Install on Zotac Zbox CI327?
by wwp
Hello there,
(after a looong time not using Fedora, just back to it!)
I've tried installing Fedora on a Zotac Zbox CI327 (Intel Celeron
N3450 inside) but this happens to fail. I've tried F25 but it has
not the right kernel to make it (I've read about 4.10 minimum), then
tried F26-Alpha-1.7 Live and Rawhide-20170518 Live but none could
reach the graphical login. They both boot, some steps take a lot of
time and I end up in a state I don't understand but which is not
friendly ;-).
Here's a link to a rdsosreport.txt file generated using Rawhide, I must
admit that I don't know what to do now..
https://paste.fedoraproject.org/paste/iRx8d~7Luk2TC3bsaDuuT15M1UNdIGYhyRL...
BTW, I could install and run a Ubuntu 17.10 daily
(artful-desktop-amd64-20170519.iso) - wow it starts up pretty quickly -
and that gave me hope!
Any hint or experience w/ such hardware?
Regards,
--
wwp
2 years, 1 month
How to turn on networking from rescue boot
by vendor@billoblog.com
So... I did a "sudo dnf upgrade" and things seemed to do well.
However, when I rebooted, it hangs on bootup with the statment "Holding
until bootup processes complete" or something like that -- I'm doing
this from memory. I tried with a couple of older kernels, but they also
hung.
I assume that this is some transient thing associated with the upgrade.
Over the years, I've had glitches appear on an upgrade and disappear on
the next upgrade. Accordingly, I'd like to boot into rescue mode and
see if I can upgrade again.
However, I've forgotten how to turn on wifi networking from rescue mode.
Can anybody point me to a tutorial?
Thanks,
billo
2 years, 2 months
F26 update: gnome-terminal doesn't launch; locale issues
by Mike Zingale
I just did an update from F25 to F26 via dnf. After it boots, I can no
longer start gnome terminal (nothing shows). If I ssh into the machine and
try it remotely, I see the following:
$ gnome-terminal
(process:3586): Gtk-WARNING **: Locale not supported by C library.
Using the fallback 'C' locale.
Error constructing proxy for
org.gnome.Terminal:/org/gnome/Terminal/Factory0: Error calling
StartServiceByName for org.gnome.Terminal: Timeout was reached
So it looks like there is a locale issue in my update. Does anyone know
how to fix this?
2 years, 8 months
how to correctly log without journald (and its blobs)?
by Franta Hanzlík
In previous Fedora distros was (IMHO) right way to log only to syslog,
without journald and its (for me) unwanted annoying binary logs, by using
configuration as:
*) Set systemd log target to syslog:
systemd.log_target=syslog (syslog-or-kmsg) ---- on kernel cmdline
systemd.default_standard_output=syslog --/
and
LogTarget=syslog ---- in systemd.conf
DefaultStandardOutput=syslog --/
*) configuring rsyslog to listen on /dev/log unix socket:
$SystemLogSocketName /dev/log ------ in rsyslog.conf
$ModLoad imuxsock ----/
$OmitLocalLogging off ---/
$AddUnixListenSocket /run/systemd/journal/syslog --/(legacy directives)
$AddUnixListenSocket /run/systemd/journal/socket -/
$AddUnixListenSocket /run/systemd/journal/stdout /
(not sure when all last three directives needed)
*) prevent to run journald:
systemctl mask systemd-journald.service
(and maybe mask systemd-journal-flush.service)
And it works fine.
But in actual Fedora 24 in systemd man page values syslog-or-kmsg and
syslog are missing in LogTarget option. As systemd/journald man pages
say hardly anything about exact mean of appropriate configuration for
this purpose, then please when someone more knowledgeable can advise:
- where to direct the systemd output? Maybe to kmsg an then read it in
rsyslog via imklog?
(rsyslog should have also imkmsg module, but is not in rsyslog-8.12.0-3.fc24)
- what about /run/log/journal/.../system.journal ?
On my test F24 system it have open rsyslogd, abrt-dump-journal-oops and
abrt-dump-journal-xorg (no need for these last two) processes. Uses it
also something else?
- what about /run/systemd/journal/{dev-log,socket,stdout} unix sockets?
Should rsyslogd listen on them?
- /dev/log seems be now symlink to /run/systemd/journal/dev-log socket
(as defined by systemd-journald-dev-log.socket) Should it be left
(because now something sends to /run/systemd/journal/dev-log), or can
be /run/systemd/journal/dev-log removed?
Thanks in advance for some clarification about this.
Franta Hanzlik
2 years, 9 months
how to (re-?)construct grub menu?
by William
Good afternoon,
I found out the rather hard way what caused those boot failure problems
that were the subject of a threat starting last month. After my weekly
"dnf upgrade" last Thursday, I checked my root e-mail, and found the
hard drive had 8 bad sectors, then 16 bad sectors. A smartctl short
test showed several parameters "pre-fail", and the others "old-age". I
obtained a new drive. My Cray Research friend gave me a USB stick with
"Clonezilla-Live-Version" on it. That duplicated the old hard drive
onto the new one, but apparently not 100%. After the old drive was
removed and the new drive installed, I could not boot. The system was
being put into a grub shell(?). (It displayed the "grub> " prompt.)
After beating out heads against the wall (or the phone), my Cray
research friend gave me another USB stick with "Boot-Repair-Disk" on it.
That rebuilt a grub file, but not correctly. But at least I got a
grub menu when booting, and I was able to edit the shell that the grub
menu entry would run. I had to take the "efi" off the commands
"linuxefi" and "initrdefi". That allows me to boot up Fedora. But the
menu is missing an entry for windows, and I still have to edit that
script every time.
I want the grub menu to offer the three most recent patches of Fedora,
the most recent Fedora rescue shell, and windows-7, in that order.
(This is a dual-boot system.) And I want the shells launched by the
menu entries to be correct. How do I get that accomplished - the
correct Fedora-25 way? If it matters, the motherboard uses UEFI bios.
Thank-you in advance.
Bill.
2 years, 10 months
Failure with Intel wireless on Dell XPS 17
by Go Canes
After a dnf update today, I was unable to bring up my normal wireless
connection on a Dell XPS 17 laptop. I was able to get a connection
via my cellphone, and did a "dnf downgrade iwl*-firmware
NetworkManager*", and after a reboot I am back online.
From info in /var/log/messages, it looks like the relevant firmware
packages are iwl7260-firmware-1:25.30.13.0-76.fc25.noarch which was
downgraded to iwl7260-firmware-1:25.30.13.0-68.fc25.noarch.
I'm happy to open a bugzilla if there isn't one on this issue, but it
seems to me I should have more info than what I posted above, and so
am looking for advice/guidance from the list.
Thanks in advance!
3 years
since upgrading to F26, 'ps' will wedge the entire OS
by Lonni J Friedman
Greetings,
I've been running Fedora since the Fedora Core days. About 11 days
ago, I upgraded my x86_64 system from 25 to 26. Prior to upgrading to
26, I was running 25 for over 6 months without any stability issues.
Since the upgrade to 26, something inspecting the process list is
wedging the OS. The problem isn't even obvious until I happen to run
'ps' and the command hangs indefinitely, and can't be killed.
However, at that point, the load avg takes off, and just keeps
climbing over time. The system is still mostly responsive, but
clearly its very unhappy about something. I'm able to check 'dmesg'
as well as /var/log/messages while its in the bad state, and there's
nothing new in either, and definitely nothing bad, unusual or
suspicious. Unfortunately, I can't even shutdown/reboot gracefully at
this point, as anything & everything that wants to inspect the process
list also hangs, and so the system never fully shuts down (I waited a
full hour the first time hoping that eventually it would proceed). I
have to hit RESET and then (thankfully) the system boots up just fine
again.
The first hang happened about 28 hours after upgrading to 26. Things
were fine after that for just over 9 days of uptime when the same
thing happened all over again.
I've tried googling to see if there's any one else running into this,
and didn't find much. This was the closest thing, but its over a year
ago:
https://forums.fedoraforum.org/archive/index.php/t-310268.html
At this point, I'm wondering whether anyone:
0) recognizes this as a known bug
1) has any ideas how to debug it further
thanks!
3 years, 1 month