looking for nonlinear optimizer with constraints accessible from
python
by stan
I want to solve a small (a few hundred variables) nonlinear optimization
problem with constraints in Fedora. I've found the scipy optimization
routines, but they are for unconstrained problems. And there are the
coin-or routines (ipopt, etc.), but they are not easily accessible from
python. Fedora doesn't package pyomo, which can access them, so I've
installed it to --user, but it seems like overkill, so I'm still
looking.
I would like something like pulp for linear programming; the closest
I've found is pyOpt, and I like its interface, but it hasn't been
maintained for 4 years, and it throws errors indicating that it might be
incompatible with more recent versions of numpy (f2py) and multiarray.
It can't find the shared libraries, compiled into python using f2py, of
the fortran77 optimization routines it uses.
So, before I go the pyomo route, does anyone have any recommendations
for simple python accessible software that would do this in Fedora?
Thanks.
6 years, 1 month
selinux question
by François Patte
Bonjour,
I have some alerts from selinux, for instance:
SELinux prevent mdadm to access getattr on file
/dev/shm/lttng-ust-wait-7-972
Why this alert? What should I do?
This is not the only one, how to configure selinux to act in a "normal"
way: leave regular processes to access what they need to work normally?
Thank you.
--
François Patte
UFR de mathématiques et informatique
Laboratoire CNRS MAP5, UMR 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)6 7892 5822
http://www.math-info.univ-paris5.fr/~patte
6 years, 1 month
PHP SOAP error after upgrade
by Chris Kottaridis
I finally got around to upgrading from Fedora25 to Fedora26.
The layout is I have two partitions of which at any time one of the is
the root partition, which contains everything but the /boot partition,
and the other is not used. I have a separate /boot partition. When I
upgrade I go into single user and duplicate the currently used root
partition to the unused one using:
# mkfs.xfs /dev/fedora/rootb
# mount /dev/fedora/rootb /mnt
# xfsdump -J - / | xfsrestore -J - /mnt
Then I modify /boot/grub2/grub.cfg and tell vmlinuz to use the new
partition and modify /mnt/etc/fstab to have the root partition be the
new partition. I then reboot and indeed I come up on the new partition
just fine, still running F25. On this new partition I do the OS upgrade.
The upgrade goes fine and it boots up just fine on Fedora26.
I did a search for .rpmsave files to see if the upgrade changed any
configuration files I may have modified locally and that came up empty.
However, when I try and use the application that runs on that machine
which is PHP SOAP I get an error "can not connect to host". The nslookup
for that host is correct and I can ping the host. As far as I know
nothing has changed on that host. My code uses the following to handle
self-signed certs:
'stream_context'=> stream_context_create(array('ssl'=>
array('verify_peer'=>false,'verify_peer_name'=>false)))
But I know that the cert on the host I am contacting is not self-signed.
If I go back and edit /boot/grub2/grub.cfg to boot up off the old
Fedora25 root partition it comes up fine and I can use the application
just fine to connect to the host.
Is there any history of PHP or SOAP having a change from F25 to F26 that
could explain the "can not connect to host" error I am seeing on F26
that I don't see on F25 ?
Thanks
Chris Kottaridis
6 years, 1 month
SELinux is blocking lightdm login to Xfce
by ToddAndMargo
Hi All,
Fedora 27, x64
Xfce 4.12
lightdm-1.25.1-5.fc27.x86_64
With SELinux set to Enforcing, I can only log into Xfce as root.
If I set SELinux to Permissive, I can log into anyone.
SEAlert is quite.
In the Audit log, I get:
# grep lightdm /var/log/audit/audit.log | grep denied
type=AVC msg=audit(1520843479.104:515): avc: denied { create } for
pid=7554 comm="lightdm" name=".xsession-errors"
scontext=system_u:system_r:xdm_t:s0-s0:c0.c1023
tcontext=system_u:object_r:samba_share_t:s0 tclass=file permissive=1
type=AVC msg=audit(1520843479.104:516): avc: denied { write open } for
pid=7554 comm="lightdm" path="/home/tony/.xsession-errors" dev="dm-1"
ino=54526689 scontext=system_u:system_r:xdm_t:s0-s0:c0.c1023
tcontext=system_u:object_r:samba_share_t:s0 tclass=file permissive=1
SELinux is taking a shine to everyone's, except root's,
.xsession-errors.
How do I fix this?
Many thanks,
-T
6 years, 1 month
RH rpms, and installing using hardlinks vs symlinks
by Robert P. J. Day
i'm curious about RH packaging policy that dictates that some
command variants are packaged for fedora to install with symlinks
and others with hardlinks.
trivial example in /usr/bin on my fedora 27 system:
-rwsr-xr-x. 1 root root 52984 Aug 2 2017 at
lrwxrwxrwx. 1 root root 2 Aug 2 2017 atq -> at
lrwxrwxrwx. 1 root root 2 Aug 2 2017 atrm -> at
so even though all of those "commands" are in the very same directory,
atq and atrm are supported via symlinks, not hardlinks.
OTOH, consider the "git" command, also in /usr/bin:
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git
as you can see, there are 116 hardlinks to that executable, pretty
much all of them in /usr/libexec/git-core:
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git-add
-rwxr-xr-x. 1 root root 41441 Feb 16 15:03 git-add--interactive
-rwxr-xr-x. 3 root root 2273360 Feb 16 15:03 git-am
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git-annotate
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git-apply
-rwxr-xr-x. 116 root root 2273360 Feb 16 15:03 git-archive
... big snip ...
is there a reason that "at" uses symlinks, while "git" installs with
hardlinks? one would imagine that it would be the other way around --
given that the "at" variations are all in the same directory,
hardlinks would seem to be the better choice since there is no
possibility of crossing filesystem boundaries.
also, what if i manually partitioned while installing, and put
/usr/bin and /usr/libexec in separate partitions? (yes, i realize
that's not the standard these days -- all of /usr defaults to a single
partition -- but i'm assuming i still have the freedom to do something
that unwise.) would git then switch to symlinks to support all those
command name variants?
just curious about how decisions like this are made WRT to
packaging.
rday
6 years, 1 month
Amd Kaveri + AMDGPU kernel driver - any success?
by Clemens Eisserer
Hi,
I would like to switch my AMD kaveri (7650k) based system from radeon
to the amdgpu kernel driver.
From what I've read, Kaveri is Southern Island and should be supported
by AMDGPU.
To force amdgpu i simply added "radeon.si_support=0
amdgpu.si_support=1" to the kernel command line,
which had the effect that my system is now hanging at the "Starting
Switch Root..." stage.
Did I miss something or isn't the amdgpu kernel module maybe not
included in fedora's initrd?
Thank you in advance, Clemens
6 years, 1 month
cups sucks!!!
by François Patte
Bonjour.
I try to configure a printer on my local network but all attemps fail.
I have a printer attached to one computer on an usb port and I want to
use it with other computers on my local network.
I opened firefox on one computer and miracle: the printer attached to
the other computer is discovered and cups admin interface asks me if I
want to add this printer. Nice: the job is easy!! *But* this does not
work at all!!
Sending a test page returns that the printer is not responding.
I tried to cofigure the printer using the "Add printer" way and chose
ipp. According to the given syntax, I wrote: ipp://name-of-server/ipp
but this does not work: sending a test page returns that the printer is
misconfigured or no longer exists...
I replaced the name of the server by its IP address... same result.
I tried: ipp://name-of-server:631/printers/name-of-printer-on-server. It
does not work, the message is now: Filter failed.
What fiter? I don't know as cups logs no longer exist since systemd...
Does somebody know a solution?
Thank you.
--
François Patte
UFR de mathématiques et informatique
Laboratoire CNRS MAP5, UMR 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)6 7892 5822
http://www.math-info.univ-paris5.fr/~patte
6 years, 1 month
how to keep NetworkManager alive after lid close ?
by sean darcy
Fedora 26. Xfce. Power Manager is set to Lock Screen when Plugged in if
laptop lid closed. And thatn works, does not suspend or hibernate. BUT,
NetworkManager is stopped. NM starts when the lid is opened.
How can I get NM to stay alive when I close the lid ?
sean
6 years, 1 month
saned and root question
by ToddAndMargo
Hi All,
Okay, now this is "scary".
Both xsane and Simple Scan work locally.
I can not get saned to work, UNLESS, I edit /etc/group
and add the following to root
root:x:0:saned
Without it, I get
$ xsane net:localhost:epkowa:interpreter:001:007
Access to resource has been denied
Now what am I doing wrong? Must saned have root privileges?
CUPS doesn't need it.
Many thanks,
-T
6 years, 1 month