Re: [gentoo-user] Re: Portage getting slicker?
On 14/07/2016 04:03, James wrote: >> It's unsurprising you got different behaviour > true, but the -u was in both and a complete different set of > packages was considered, by portage, and only one was able to > move forward (note the -p was not in the second entry, despite > my not including that detail). without -N or -t, portage considers just the list of packages on the command line. -N is newuse, portage also considers packages whose USE has changed. -t is emptytree, portage also considers the entire tree and -u tells it to not remerge things that don't need updating. The input set for those commands differs, so the output set might also be different. Those two commands you ran are not guaranteed to produce the same results (although the often will). -- Alan McKinnon alan.mckin...@gmail.com
[gentoo-user] Re: Portage getting slicker?
Alan McKinnon gmail.com> writes: > > Ah excellent point, but the build did not move forward with:: > > ' emerge -uvDN world' either. With the --tree it did move forward with > > the build update. In the first attempt usually the packages to be built > > are listed, conflicts or blockers. > But you didn't run > emerge -uvDN world > You ran > emerge -uvDNp world > why won't move forward, ever Nope. I ran 'emerge -uvDNp world' and then 'emerge -uvDN world' No point delineating that detail, or so I thought > > None of these 3 packages where listed in the first attempt to see > > what needs to be built:: > > Not 'sys-devel/llvm', nor 'sys-devel/clang', nor 'media-libs/mesa'. > >> Emerging (1 of 3) sys-devel/llvm-3.7.1-r3::gentoo > >>> I did nothing manual in between. Explanations? > >> portage is doing what's expected. You don't have -a in the command line > >> and there's nothing stopping portage from moving forward with the build. > >> SO it moved forward with the build. > > > > Yes, nothing to do with 'media-libs/jasper' nor 'gdk-pixbuf'. So I guess the > > --tree option got rid of the these (conflicts issues. My point is that this > > is remarkably better than how things worked in the past (but not certain > > when these enhancements were made). > > But you introduced two significant changes in you command line > removed -N > added -t > It's unsurprising you got different behaviour true, but the -u was in both and a complete different set of packages was considered, by portage, and only one was able to move forward (note the -p was not in the second entry, despite my not including that detail).
Re: [gentoo-user] Re: Portage getting slicker?
On 13/07/2016 23:41, James wrote: Alan McKinnon gmail.com> writes: On 13/07/2016 20:25, James wrote: So, today I ran a sync and upgrade to a gentoo workstation:: emerge -uvDNp world These are the packages that would be merged, in order: Calculating dependencies... done! Total: 0 packages, Size of downloads: 0 KiB WARNING: One or more updates/rebuilds have been skipped due to a dependency conflict: media-libs/jasper:0 (media-libs/jasper-1.900.1-r9:0/0::gentoo, ebuild scheduled for merge) conflicts with media-libs/jasper:0/0=[abi_x86_32(-),abi_x86_64(-)] required by (x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed) media-libs/jasper:=[abi_x86_32(-),abi_x86_64(-)] required by (x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed) This is not a blocker. Read the warning, it says an update or rebuild was skipped due to a dependency conflict. In your casejasper-1.900.1-r9 was not done due to gdk-pixbuf requirements. Presumably, what you already have keeps pixbuf happy Blockers in that output usually have "!!" annotations at the beginning. Ah excellent point, but the build did not move forward with:: ' emerge -uvDN world' either. With the --tree it did move forward with the build update. In the first attempt usually the packages to be built are listed, conflicts or blockers. But you didn't run emerge -uvDN world You ran emerge -uvDNp world why won't move forward, ever None of these 3 packages where listed in the first attempt to see what needs to be built:: Not 'sys-devel/llvm', nor 'sys-devel/clang', nor 'media-libs/mesa'. Emerging (1 of 3) sys-devel/llvm-3.7.1-r3::gentoo I did nothing manual in between. Explanations? portage is doing what's expected. You don't have -a in the command line and there's nothing stopping portage from moving forward with the build. SO it moved forward with the build. Yes, nothing to do with 'media-libs/jasper' nor 'gdk-pixbuf'. So I guess the --tree option got rid of the these (conflicts issues. My point is that this is remarkably better than how things worked in the past (but not certain when these enhancements were made). But you introduced two significant changes in you command line removed -N added -t It's unsurprising you got different behaviour
[gentoo-user] Re: Is "-fomit-frame-pointer" a gcc default?
On Wed, 13 Jul 2016 16:53:40 -0400 waltd...@waltdnes.org wrote: > On Tue, Jul 12, 2016 at 05:09:28PM -0500, »Q« wrote > > On Tue, 12 Jul 2016 12:14:57 -0500 > > R0b0t1 wrote: > > > > > Pale Moon is routinely behind Firefox on security fixes (actual > > > fixes, not wanking-in-a-corner fixes). > > > > Is anyone other than the Pale Moon team itself trying to track its > > vulnerabilities? I could only find one CVE for it, from 2013. > > See http://www.palemoon.org/releasenotes.shtml with several mentions > of CVEs and other security fixes. Given the amount of Firefox code > still present "under the hood", many Firefox security fixes will also > apply to Pale Moon. Checking just a few, the Pale Moon team takes anywhere from a few weeks to a few months to fix security vulnerabilities which have been published and fixed by Mozilla. And other Firefox CVEs aren't listed by Pale Moon, so it's tough to tell whether or not Pale Moon is/was affected. Maybe their fork of Gecko has diverged too much to easily port Mozilla's fixes, I dunno. But not to worry, they have a FAQ. Is Pale Moon safe to use? Absolutely! Pale Moon is based on the Mozilla release source code that has a large community of developers and security-aware people, next to having seen over a decade of development by now. [...] OTOH, when it suits him, Moonchild stresses how very different his codebase is now from Mozilla's. AFAICS, no one but the Pale Moon team is tracking Pale Moon vulnerabilities. I dunno what to make of their claims that it's safe to use.
[gentoo-user] Re: Gitlab experiences
Alan McKinnon gmail.com> writes: > > often a custom, scheduler/framework has to be modified or custom developed, > > but I find it hard to believe there is no way to massively speed up > > something like gitlab on a robust linux cluster. That dragon has been > > sleighed for most all load problems, be it HPC, networking issues, or > > security audits, etc etc. Often, a cloud/cluster software can be radically > > sped up by allocating lots of extra ram to the framework it is running on. > > Cloud vendors charge way to much for extra ram, so performance of > > ram-intensive services are often run where there is ample ram. > Depends. It might all be running on a single-threaded ruby process in > the middleware If you believe what one reads:: "You seem to be confusing two very different things here: the Ruby Programming Language and the specific threading model of one specific implementation of the Ruby Programming Language. There are currently around 11 different implementations of the Ruby Programming Language, with very different and unique threading models. " circa 2011. > Don't laugh, I've seen stranger things. > Alan There are a multitude of ways to profile codes and the modules. Surely someone at gitlab has done this? (Evidently No?) A big part of cluster architecture is to identify the bottlenecks in codes that run on a single machine and implement solutions to lessen the bottlenecks. Most modern codes inherently run parallel and that's what cluster provide, more resources for many more threads, and the addition of custom hardware, such as GPUs, DDR5 ram, fpga and different architectures of processors. Surely gitlab can be fixed with a correctly applied cluster, or the codes that compose gitlab, need to be fixed... I'll be the first one to verify, you just don't rent a cluster (containers/VM) and move you slow applications there and magically all is fixed. The first step is to identify and qualify/quantify the source(s) of the bottlnecks. CS-101. However, profiling code/kernels/libs often leads one down the 'fuzzer' path, and that's a whole other (fun) time-sink. ymmv, James
Re: [gentoo-user] Re: qtcore 5 missing the lrelease binary?
On Wed, 13 Jul 2016 23:24:35 +0200, Alan McKinnon wrote: > > Instant help from Alan and Neil, two foreigners from the other side of > > the planet. Thanks, guys :) I'll let you know next week when I > > finish rebuilding the machine with every permutation of useflags I > > can think of. > s/foreigners/cranky old foreign farts/g > > There ya go, fixed it for you. > > [Neil is going to argue those adjectives but they work for me ] :-) I was going to argue the foreign part. You and Walt are foreigners, I'm not! -- Neil Bothwick "Be strict when sending and tolerant when receiving." RFC 1958 - Architectural Principles of the Internet - section 3.9 pgpnxQb54SIYc.pgp Description: OpenPGP digital signature
[gentoo-user] Re: Portage getting slicker?
Alan McKinnon gmail.com> writes: > > On 13/07/2016 20:25, James wrote: > >> > > So, today I ran a sync and upgrade to a gentoo workstation:: > > emerge -uvDNp world > > > > These are the packages that would be merged, in order: > > Calculating dependencies... done! > > Total: 0 packages, Size of downloads: 0 KiB > > WARNING: One or more updates/rebuilds have been skipped due to a dependency > > conflict: > > > > media-libs/jasper:0 > > > >(media-libs/jasper-1.900.1-r9:0/0::gentoo, ebuild scheduled for merge) > > conflicts with > > media-libs/jasper:0/0=[abi_x86_32(-),abi_x86_64(-)] required by > > (x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed) > > > > media-libs/jasper:=[abi_x86_32(-),abi_x86_64(-)] required by > > (x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed) > > This is not a blocker. > > Read the warning, it says an update or rebuild was skipped due to a > dependency conflict. In your casejasper-1.900.1-r9 was not done due to > gdk-pixbuf requirements. Presumably, what you already have keeps pixbuf > happy > > Blockers in that output usually have "!!" annotations at the beginning. Ah excellent point, but the build did not move forward with:: ' emerge -uvDN world' either. With the --tree it did move forward with the build update. In the first attempt usually the packages to be built are listed, conflicts or blockers. None of these 3 packages where listed in the first attempt to see what needs to be built:: Not 'sys-devel/llvm', nor 'sys-devel/clang', nor 'media-libs/mesa'. > Emerging (1 of 3) sys-devel/llvm-3.7.1-r3::gentoo > > I did nothing manual in between. Explanations? > portage is doing what's expected. You don't have -a in the command line > and there's nothing stopping portage from moving forward with the build. > SO it moved forward with the build. Yes, nothing to do with 'media-libs/jasper' nor 'gdk-pixbuf'. So I guess the --tree option got rid of the these (conflicts issues. My point is that this is remarkably better than how things worked in the past (but not certain when these enhancements were made). Thanks for pointing out blockers vs conflicts... James
Re: [gentoo-user] Re: qtcore 5 missing the lrelease binary?
On 13/07/2016 23:21, walt wrote: On Wed, 13 Jul 2016 21:42:37 +0100 Neil Bothwick wrote: On Wed, 13 Jul 2016 13:35:26 -0700, walt wrote: I just installed qt5 for the first time because the latest virtualbox requires it. Now virtualbox 5.1.0 won't build because it can't find /usr/lib64/qt5/bin/lrelease, which (I think) should be in the qtcore package. % qfile /usr/lib64/qt5/bin/lrelease dev-qt/linguist-tools (/usr/lib64/qt5/bin/lrelease) Instant help from Alan and Neil, two foreigners from the other side of the planet. Thanks, guys :) I'll let you know next week when I finish rebuilding the machine with every permutation of useflags I can think of. s/foreigners/cranky old foreign farts/g There ya go, fixed it for you. [Neil is going to argue those adjectives but they work for me ] :-)
[gentoo-user] Re: qtcore 5 missing the lrelease binary?
On Wed, 13 Jul 2016 21:42:37 +0100 Neil Bothwick wrote: > On Wed, 13 Jul 2016 13:35:26 -0700, walt wrote: > > > I just installed qt5 for the first time because the latest > > virtualbox requires it. Now virtualbox 5.1.0 won't build because > > it can't find /usr/lib64/qt5/bin/lrelease, which (I think) should > > be in the qtcore package. > > % qfile /usr/lib64/qt5/bin/lrelease > dev-qt/linguist-tools (/usr/lib64/qt5/bin/lrelease) Instant help from Alan and Neil, two foreigners from the other side of the planet. Thanks, guys :) I'll let you know next week when I finish rebuilding the machine with every permutation of useflags I can think of.
Re: [gentoo-user] Portage getting slicker?
On 13/07/2016 20:25, James wrote:
So, today I ran a sync and upgrade to a gentoo workstation::
emerge -uvDNp world
These are the packages that would be merged, in order:
Calculating dependencies... done!
Total: 0 packages, Size of downloads: 0 KiB
WARNING: One or more updates/rebuilds have been skipped due to a dependency
conflict:
media-libs/jasper:0
(media-libs/jasper-1.900.1-r9:0/0::gentoo, ebuild scheduled for merge)
conflicts with
media-libs/jasper:0/0=[abi_x86_32(-),abi_x86_64(-)] required by
(x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed)
media-libs/jasper:=[abi_x86_32(-),abi_x86_64(-)] required by
(x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed)
This is not a blocker.
Read the warning, it says an update or rebuild was skipped due to a
dependency conflict. In your casejasper-1.900.1-r9 was not done due to
gdk-pixbuf requirements. Presumably, what you already have keeps pixbuf
happy
Blockers in that output usually have "!!" annotations at the beginning.
OK, no big deal, except I next ran::
'emerge -uDtv @world'
BANG, it just started compiling as if no blokers?
Calculating dependencies... done!
[ebuild rR] media-libs/mesa-11.0.6::gentoo USE="classic dri3 egl
gallium gbm gles2 llvm nptl openmax udev vaapi vdpau xvmc -bindist -d3d9
-debug -gles1 -opencl -osmesa -pax_kernel -pic (-selinux) -wayland -xa"
ABI_X86="32 (64) (-x32)" VIDEO_CARDS="radeon radeonsi (-freedreno) -i915
-i965 -ilo -intel -nouveau -r100 -r200 -r300 -r600 -vmware" 7,103 KiB
[nomerge ] sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo
[3.5.0:0/3.5::gentoo] USE="clang libffi ncurses python static-analyzer
-debug -doc -gold -libedit -lldb% -multitarget -ocaml {-test} -xml"
ABI_X86="32 (64) (-x32)" PYTHON_TARGETS="python2_7 (-pypy%)"
VIDEO_CARDS="radeon"
[blocks b ] <=sys-devel/clang-3.7.1-r99 ("<=sys-devel/clang-3.7.1-r99"
is blocking sys-devel/llvm-3.7.1-r3)
[ebuild U ] sys-devel/clang-3.7.1-r100:0/3.7::gentoo
[3.5.0-r100:0/3.5::gentoo] USE="python static-analyzer -debug -multitarget"
ABI_X86="(64) -32 (-x32)" 0 KiB
[ebuild r U ]sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo
[3.5.0:0/3.5::gentoo] USE="clang libffi ncurses python static-analyzer
-debug -doc -gold -libedit -lldb% -multitarget -ocaml {-test} -xml"
ABI_X86="32 (64) (-x32)" PYTHON_TARGETS="python2_7 (-pypy%)"
VIDEO_CARDS="radeon" 24,610 KiB
Total: 3 packages (2 upgrades, 1 reinstall), Size of downloads: 31,713 KiB
Conflict: 1 block
The following packages are causing rebuilds:
(sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo, ebuild scheduled for merge)
causes rebuilds for:
(media-libs/mesa-11.0.6:0/0::gentoo, ebuild scheduled for merge)
Verifying ebuild manifests
Running pre-merge checks for sys-devel/llvm-3.7.1-r3
* Checking for at least 2400 MiB disk space at
"/var/tmp/portage/sys-devel/llvm-3.7.1-r3/temp" ...
[ ok ]
Emerging (1 of 3) sys-devel/llvm-3.7.1-r3::gentoo
Now this is col
I did nothing manual in between. Explanations?
portage is doing what's expected. You don't have -a in the command line
and there's nothing stopping portage from moving forward with the build.
SO it moved forward with the build.
Re: [gentoo-user] Re: Gitlab experiences
On 13/07/2016 21:13, James wrote: Jeremi Piotrowski gmail.com> writes: Has anyone attempted to install a self hosted gitlab on gentoo server(s)? I would deploy it with docker. The gitlab guys push official images of the main gitlab app[1] and CI runners[2] to dockerhub. That should be the easiest path to getting it up and running in no time. Docker runs everywhere, including Mesos clusters. That being said, gitlab does not really play well with clustering in general. I don't think the main part of the app does any kind of horizontal scaling (gitlab.com is hosted on a single server) so you need a fairly beefy server. That's weird. Almost every type of heavy load is finding it's way to clusters now; many via containers some on bare metal clusters. Granted, very often a custom, scheduler/framework has to be modified or custom developed, but I find it hard to believe there is no way to massively speed up something like gitlab on a robust linux cluster. That dragon has been sleighed for most all load problems, be it HPC, networking issues, or security audits, etc etc. Often, a cloud/cluster software can be radically sped up by allocating lots of extra ram to the framework it is running on. Cloud vendors charge way to much for extra ram, so performance of ram-intensive services are often run where there is ample ram. Depends. It might all be running on a single-threaded ruby process in the middleware :-) Don't laugh, I've seen stranger things. Alan
Re: [gentoo-user] Re: Gitlab experiences
On 13/07/2016 20:14, James wrote: "J." García gmail.com> writes: I know the Gentoo Infra team has had negative experiences with hosting just about anything Java and don't want to go near it. I don't know if that is based on specific experiences with GitLab or with just avoidance with Java in general. Most of the competing solutions in this space are also Java-based which is why we don't host any kind of alternative to Github. What java has to do with gitlab? according to the repo I see is mostly ruby code[1](both gitlab and gitlab-ci). what you wrote make it seems like it is a java app. In the github mirror of gitlab(the main app), the file stats are: 2,253 Ruby 697 Haml 319 Markdown 158 CoffeeScript 99 SCSS 90 Cucumber 40 YAML 39 HTML+ERB 26 SVG 25 JavaScript Huh. (Double huh...) As Alan is always quick to quip:: JavaScript ain't java, so that changes the entire game. Performance can be fixed with a gentoo cluster (yet to be proven). I also read that gitlab is working to make the "engine" faster. I have two gitlabs at work The old one which has been broken for yonks since an up grade that went terribly bad. New keys for users always made it into the gitlab db, never into git's ~/.ssh/authorized_keys, we have a handy script to restart the whole damn daemon because it hiccups every other day. Ever since a bad upgrade. The new one sings along, and I moved all my ansible stuff to it. That interne did a good job with that projects. But gitlab isn't Java in any way, don't know why Rich said that - he may have conflated gitlab with some other git*. Gitlab is ruby. With this app, and actually all webapps running in interpreted frameworks, I support Rich's idea of running a prebuilt image in a container as an appliance. Just mount your storage into the appliance, and keep the database somewhere else. I've had too many screwups with updates - that ecosystem doesn't seem to care much for upgrade paths other than to make sure the dev could git pull daily and it would continue to work commit by commit :-) Alan
Re: [gentoo-user] Re: Is "-fomit-frame-pointer" a gcc default?
On Tue, Jul 12, 2016 at 05:09:28PM -0500, »Q« wrote > On Tue, 12 Jul 2016 12:14:57 -0500 > R0b0t1 wrote: > > > Pale Moon is routinely behind Firefox on security fixes (actual fixes, > > not wanking-in-a-corner fixes). > > Is anyone other than the Pale Moon team itself trying to track its > vulnerabilities? I could only find one CVE for it, from 2013. See http://www.palemoon.org/releasenotes.shtml with several mentions of CVEs and other security fixes. Given the amount of Firefox code still present "under the hood", many Firefox security fixes will also apply to Pale Moon. -- Walter Dnes I don't run "desktop environments"; I run useful applications
Re: [gentoo-user] qtcore 5 missing the lrelease binary?
On 13/07/2016 22:35, walt wrote: I just installed qt5 for the first time because the latest virtualbox requires it. Now virtualbox 5.1.0 won't build because it can't find /usr/lib64/qt5/bin/lrelease, which (I think) should be in the qtcore package. I copied lrelease from qtcore-4, which worked around that problem but now vbox fails to build because of an undefined GL variable, so I've got other (useflag?) problems to solve, but meanwhile I'd be grateful for any hints. Thanks. It comes from linguist-tools, looks like you'll get it with USE="nls" (or any of several USE cases) alan@khamul ~ $ equery belongs /usr/lib64/qt5/bin/lrelease * Searching for /usr/lib64/qt5/bin/lrelease ... dev-qt/linguist-tools-5.6.1 (/usr/lib64/qt5/bin/lrelease) alan@khamul ~ $ equery depends linguist-tools * These packages depend on linguist-tools: app-text/sigil-0.9.5-r1 (>=dev-qt/linguist-tools-5.4:5) dev-db/sqlitebrowser-3.8.0 (qt5 ? dev-qt/linguist-tools:5) kde-apps/kde-l10n-16.04.2 (>=dev-qt/linguist-tools-5.5.1:5) kde-apps/ktp-l10n-16.04.2 (>=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/extra-cmake-modules-5.24.0 (test ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kauth-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kbookmarks-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kcodecs-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kcompletion-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kconfig-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kcoreaddons-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kcrash-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kdbusaddons-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kdesignerplugin-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kdnssd-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kglobalaccel-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kitemviews-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kjobwidgets-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/knotifications-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kwidgetsaddons-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/kwindowsystem-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/solid-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) kde-frameworks/sonnet-5.24.0 (nls ? >=dev-qt/linguist-tools-5.5.1:5) x11-misc/sddm-0.13.0-r4 (dev-qt/linguist-tools:5) Alan
Re: [gentoo-user] qtcore 5 missing the lrelease binary?
On Wed, 13 Jul 2016 13:35:26 -0700, walt wrote: > I just installed qt5 for the first time because the latest virtualbox > requires it. Now virtualbox 5.1.0 won't build because it can't > find /usr/lib64/qt5/bin/lrelease, which (I think) should be in the > qtcore package. % qfile /usr/lib64/qt5/bin/lrelease dev-qt/linguist-tools (/usr/lib64/qt5/bin/lrelease) -- Neil Bothwick SEX ON TV HAS TO GO! I keep falling off! pgp1ndHDRYfhM.pgp Description: OpenPGP digital signature
Re: [gentoo-user] executing a command as a nologin user
On 13/07/2016 19:05, wabe wrote: Alan McKinnon wrote: On 13/07/2016 18:42, wabe wrote: Fernando Rodriguez wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 07/13/2016 07:10 AM, Alan McKinnon wrote: On 12/07/2016 03:47, jens w wrote: .procmailrc :0 c * !^X-Loop: n...@example.com | formail -X "From:" | $HOME/bin/script.sh procmail.log procmail: Executing " formail -X "From:" | $HOME/bin/script.sh for incoming mail, a script is executed. logfile has the same entry as it is in other users. but the script do nothing. How executing a command as a nologin user? You can't, not the way you are doing it. You want to launch a shell script for the user, but the user's shell is /sbin/nologin. This exits immediately without launching the script. Give the user a real shell. Alan I've been following this thread and thinking the same thing but wasn't sure. I don't think so. To proof it, I created this user: nologinuser:x:1015:1016::/home/nologinuser:/sbin/nologin Then I created this script: #!/bin/sh # date >> /home/nologinuser/envars.txt echo $HOME >> /home/nologinuser/envars.txt echo $PATH >> /home/nologinuser/envars.txt echo "---" >> /home/nologinuser/envars.txt I stored it as /var/script-nologinuser/testscript.sh. I had to store it at this place because /home is mounted with the noexec option on my system. Then I created a cronjob for the nologinuser user: */2 * * * * /var/script-nologinuser/testscript.sh The result is the file /home/nologinuser/envars.txt with this content: Wed Jul 13 18:10:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- Wed Jul 13 18:12:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- The ownership and the rights for /home/nologinuser/, /var/script-nologinuser/ and /var/script-nologinuser/testscript.sh are nologinuser:nologinuser and 700. So it seems, that it is possible to execute scripts without setting a shell in /etc/passwd. I don't know why it doesn't work for jens w. Not so. Your script is launched by cron, running as root. It starts a non-interactive no-login shell (that's why people have infernal trouble with cron, assuming it has a $PATH when it actually has none) Ok, this seems to make sense. I did another test to make sure that it works also when the script isn't executed directly by crontab. I'm not sure if this is a real proof, because the "execution chain" is started by cron. I renamed /var/script-nologinuser/testscript.sh to /var/script-nologinuser/testscript-2.sh and created a script /var/script-nologinuser/testscript.sh with this content: #!/bin/sh # /var/script-nologinuser/testscript-2.sh It also worked. I think in that circumstance, cron forks and execs an sh process in a non-interactive non-login shell, so it works out OK. Now to get procmail to do the same Alan
[gentoo-user] qtcore 5 missing the lrelease binary?
I just installed qt5 for the first time because the latest virtualbox requires it. Now virtualbox 5.1.0 won't build because it can't find /usr/lib64/qt5/bin/lrelease, which (I think) should be in the qtcore package. I copied lrelease from qtcore-4, which worked around that problem but now vbox fails to build because of an undefined GL variable, so I've got other (useflag?) problems to solve, but meanwhile I'd be grateful for any hints. Thanks.
Re: [gentoo-user] no sound through headphones (OK through speakers)
On Wed, Jul 13 2016, wabe wrote: > allan gottlieb wrote: > >> On Wed, Jul 13 2016, wabe wrote: >> >> > allan gottlieb wrote: >> > >> >> On Wed, Jul 13 2016, Canek Peláez Valdés wrote: >> >> >> >> > On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb >> >> > wrote: >> >> >> Hardware: Dell Latitude E7450 laptop >> >> >> >> >> >> Gentoo essentially all stable >> >> >> Gnome / Systemd >> >> >> >> >> >> When I play a movie using totem it sounds fine if no headphones >> >> >> are plugged it. >> >> >> >> >> >> No sound at all with headphones (I tried three different ones). >> >> >> >> >> >> The sound settings gui recognizes that headphones are in. The >> >> >> sound test is silent. If I select the internal speakers in the >> >> >> gui (with the headphones still in) sound is fine. >> >> >> >> >> >> Is there some headphone option I must enable in the kernel or >> >> >> elsewhere? >> >> > >> >> > The volume is probably muted for the headphones. Install >> >> > pavucontrol, and execute it while the movie is playing. In the >> >> > "Output Devices" tab look for your sound card (probably something >> >> > like "Built-in Audio"), and in port select "Headphones". Then >> >> > adjust the volume. >> >> > >> >> > Regards. >> >> >> >> Thank you (and robot1). I installed pavucontrol and followed your >> >> instructions. The volume was selected in the middle (100%). We >> >> see a volume meter going up and down as expected with either >> >> headphones or speakers selected. When mute is pressed the meter >> >> looks dead, as expected. >> >> >> >> It also shows line-out is unplugged and headphones are plugged it. >> >> >> >> Any thoughts. I appreciate the help. >> > >> > Maybe the headphone amp or the headphone socket of your soundcard >> > is broken. Plug the speakers into the headphone socket and test if >> > you can hear something. For my experience you can do that without >> > the risk of killing something, but of course I can give you no >> > guarantee. :-) >> e> >> > -- >> > Regards >> > wabe >> >> I can't do that this week. I can try it next week at home where I >> think I have speakers that plug in. > > If your headphones are headsets with microphones it maybe could be > that the headphone socket of the soundcard isn't compatible with > the plugs of the headphones. But that's just a thought. I never had > a soundcard where this was the case. > > -- > Regards > wabe They are just earbuds, no mic. But maybe the port of the laptop assumes a mic + headphone (i.e., three signals mic, left "ear", right "ear" as used for cell phones. I have one of those at home I can try this weekend. thanks, allan
Re: [gentoo-user] no sound through headphones (OK through speakers)
allan gottlieb wrote: > On Wed, Jul 13 2016, wabe wrote: > > > allan gottlieb wrote: > > > >> On Wed, Jul 13 2016, Canek Peláez Valdés wrote: > >> > >> > On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb > >> > wrote: > >> >> Hardware: Dell Latitude E7450 laptop > >> >> > >> >> Gentoo essentially all stable > >> >> Gnome / Systemd > >> >> > >> >> When I play a movie using totem it sounds fine if no headphones > >> >> are plugged it. > >> >> > >> >> No sound at all with headphones (I tried three different ones). > >> >> > >> >> The sound settings gui recognizes that headphones are in. The > >> >> sound test is silent. If I select the internal speakers in the > >> >> gui (with the headphones still in) sound is fine. > >> >> > >> >> Is there some headphone option I must enable in the kernel or > >> >> elsewhere? > >> > > >> > The volume is probably muted for the headphones. Install > >> > pavucontrol, and execute it while the movie is playing. In the > >> > "Output Devices" tab look for your sound card (probably something > >> > like "Built-in Audio"), and in port select "Headphones". Then > >> > adjust the volume. > >> > > >> > Regards. > >> > >> Thank you (and robot1). I installed pavucontrol and followed your > >> instructions. The volume was selected in the middle (100%). We > >> see a volume meter going up and down as expected with either > >> headphones or speakers selected. When mute is pressed the meter > >> looks dead, as expected. > >> > >> It also shows line-out is unplugged and headphones are plugged it. > >> > >> Any thoughts. I appreciate the help. > > > > Maybe the headphone amp or the headphone socket of your soundcard > > is broken. Plug the speakers into the headphone socket and test if > > you can hear something. For my experience you can do that without > > the risk of killing something, but of course I can give you no > > guarantee. :-) > e> > > -- > > Regards > > wabe > > I can't do that this week. I can try it next week at home where I > think I have speakers that plug in. If your headphones are headsets with microphones it maybe could be that the headphone socket of the soundcard isn't compatible with the plugs of the headphones. But that's just a thought. I never had a soundcard where this was the case. -- Regards wabe
Re: [gentoo-user] no sound through headphones (OK through speakers)
On Wed, Jul 13 2016, wabe wrote: > allan gottlieb wrote: > >> On Wed, Jul 13 2016, Canek Peláez Valdés wrote: >> >> > On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb >> > wrote: >> >> Hardware: Dell Latitude E7450 laptop >> >> >> >> Gentoo essentially all stable >> >> Gnome / Systemd >> >> >> >> When I play a movie using totem it sounds fine if no headphones are >> >> plugged it. >> >> >> >> No sound at all with headphones (I tried three different ones). >> >> >> >> The sound settings gui recognizes that headphones are in. The >> >> sound test is silent. If I select the internal speakers in the >> >> gui (with the headphones still in) sound is fine. >> >> >> >> Is there some headphone option I must enable in the kernel or >> >> elsewhere? >> > >> > The volume is probably muted for the headphones. Install >> > pavucontrol, and execute it while the movie is playing. In the >> > "Output Devices" tab look for your sound card (probably something >> > like "Built-in Audio"), and in port select "Headphones". Then >> > adjust the volume. >> > >> > Regards. >> >> Thank you (and robot1). I installed pavucontrol and followed your >> instructions. The volume was selected in the middle (100%). We see a >> volume meter going up and down as expected with either headphones or >> speakers selected. When mute is pressed the meter looks dead, as >> expected. >> >> It also shows line-out is unplugged and headphones are plugged it. >> >> Any thoughts. I appreciate the help. > > Maybe the headphone amp or the headphone socket of your soundcard is > broken. Plug the speakers into the headphone socket and test if you > can hear something. For my experience you can do that without the risk > of killing something, but of course I can give you no guarantee. :-) e> > -- > Regards > wabe I can't do that this week. I can try it next week at home where I think I have speakers that plug in. allan
Re: [gentoo-user] no sound through headphones (OK through speakers)
allan gottlieb wrote: > On Wed, Jul 13 2016, Canek Peláez Valdés wrote: > > > On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb > > wrote: > >> Hardware: Dell Latitude E7450 laptop > >> > >> Gentoo essentially all stable > >> Gnome / Systemd > >> > >> When I play a movie using totem it sounds fine if no headphones are > >> plugged it. > >> > >> No sound at all with headphones (I tried three different ones). > >> > >> The sound settings gui recognizes that headphones are in. The > >> sound test is silent. If I select the internal speakers in the > >> gui (with the headphones still in) sound is fine. > >> > >> Is there some headphone option I must enable in the kernel or > >> elsewhere? > > > > The volume is probably muted for the headphones. Install > > pavucontrol, and execute it while the movie is playing. In the > > "Output Devices" tab look for your sound card (probably something > > like "Built-in Audio"), and in port select "Headphones". Then > > adjust the volume. > > > > Regards. > > Thank you (and robot1). I installed pavucontrol and followed your > instructions. The volume was selected in the middle (100%). We see a > volume meter going up and down as expected with either headphones or > speakers selected. When mute is pressed the meter looks dead, as > expected. > > It also shows line-out is unplugged and headphones are plugged it. > > Any thoughts. I appreciate the help. Maybe the headphone amp or the headphone socket of your soundcard is broken. Plug the speakers into the headphone socket and test if you can hear something. For my experience you can do that without the risk of killing something, but of course I can give you no guarantee. :-) -- Regards wabe
Re: [gentoo-user] no sound through headphones (OK through speakers)
On Wed, Jul 13 2016, Canek Peláez Valdés wrote: > On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb wrote: >> Hardware: Dell Latitude E7450 laptop >> >> Gentoo essentially all stable >> Gnome / Systemd >> >> When I play a movie using totem it sounds fine if no headphones are >> plugged it. >> >> No sound at all with headphones (I tried three different ones). >> >> The sound settings gui recognizes that headphones are in. The sound >> test is silent. If I select the internal speakers in the gui (with the >> headphones still in) sound is fine. >> >> Is there some headphone option I must enable in the kernel or elsewhere? > > The volume is probably muted for the headphones. Install pavucontrol, > and execute it while the movie is playing. In the "Output Devices" tab > look for your sound card (probably something like "Built-in Audio"), > and in port select "Headphones". Then adjust the volume. > > Regards. Thank you (and robot1). I installed pavucontrol and followed your instructions. The volume was selected in the middle (100%). We see a volume meter going up and down as expected with either headphones or speakers selected. When mute is pressed the meter looks dead, as expected. It also shows line-out is unplugged and headphones are plugged it. Any thoughts. I appreciate the help. allan
[gentoo-user] Re: Gitlab experiences
Jeremi Piotrowski gmail.com> writes: > > > Has anyone attempted to install a self hosted gitlab on gentoo server(s)? > I would deploy it with docker. The gitlab guys push official images of the > main gitlab app[1] and CI runners[2] to dockerhub. That should be > the easiest path to getting it up and running in no time. Docker runs everywhere, including Mesos clusters. > That being said, gitlab does not really play well with clustering in > general. I don't think the main part of the app does any kind of > horizontal scaling (gitlab.com is hosted on a single server) so you need a > fairly beefy server. That's weird. Almost every type of heavy load is finding it's way to clusters now; many via containers some on bare metal clusters. Granted, very often a custom, scheduler/framework has to be modified or custom developed, but I find it hard to believe there is no way to massively speed up something like gitlab on a robust linux cluster. That dragon has been sleighed for most all load problems, be it HPC, networking issues, or security audits, etc etc. Often, a cloud/cluster software can be radically sped up by allocating lots of extra ram to the framework it is running on. Cloud vendors charge way to much for extra ram, so performance of ram-intensive services are often run where there is ample ram. > And while storage should be entirely up to you (the > app _should_ be indifferent to what you use) most folks appear to run with > local disks or NFS. NFS is yesterday's cruft. Many are now using cephfs to replace NFS. Surely a distributed file system (DFS) to work with the (master/slave) nodes in a cluster is also keenly important. Granted the daddy of all cluster file systems (HDFS) is a pig, but it is being actively replaces with a myriad of DFS solutions. Using NFS in a linux cluster turns performance into a cluster*F!!!. Just say no to NFS and learn about DFS. NFS is what clusters used 15 years ago (beowulf/pvm). OrangeFS is an outgrowth if the old PVM and is now part of the kernel (4.6+). Any and all references as to why gitlab crawls, the more technical the better, are of keen interest to me. It's not that I do not believe what you are saying, it just more likely that someone with keen cluster skills has not looked at gitlab. ymmv. The fact of what you are saying that gitlab does not run fast on a cluster, is very bewildering to me, and wreaks of poor architecture on the cluster it was tested on, more than anything else. Scheduler/config/frameworks are properly needed to get large loads to fly on a cluster. At this point, it a lot of work and a bit of black-arts Folks with deep pockets, are not failing to make any workload fly, in my extensive research experiences. > > > Any experiences with gitlab are most welcome for comment, good or bad. > > Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with > > large repos (except you have unlimited access to hardware resources). A > > Linux kernel git mirror finally crashed it. What are the specifics of the mirror setup, DFS and the crash? > I can second that - it's slow and load times are long even when performing > the most basic operations such as opening a small file or viewing a single > commit (you _will_ notice - it regulary takes several seconds). Mind you > the gitlab folks are working on improving this since several releases. Sounds like they need a cluster architect to help them select the relevant cluster tools and tune/modify/develop a custom scheduler and frameworks, for each of the sub problems. Also, most projects are fairly clueless about tuning the underlying linux kernels for cluster operations. Stock vendor kernels are 'dog-crap' for linux clusters, but that is another entire set of problems. > As a user, another issue I have with it is that the merge request/review > interface is just terrible. There is _no_ merge request versioning, so > either you submit your code in perfect shape at first try, or any change > (amending/rebasing/merging) will cause the changelist to be duplicated > many times over. You also lose track of the review history instantly - old > comments are either concealed or swallowed. > There is also no CLI utility to automate common review-related tasks > (submitting/responding to review) so you are forced to do everything over > the slow WebUI. That does sound bad... > If you care at all about the codereview aspects, I would recommend gerrit > or phabricator. Both have cli utilities (git-review and arcanist) and > while some claim they are ugly (heard that one especially about > gerrit) they are 100x more practical. > If you only care about having a repository browser then gitlab can work > but there are simpler apps out there (gogs/pagure). Interesting comments. thanks, James
Re: [gentoo-user] KDE 5
On Wed, Jul 13, 2016 at 11:51:28AM +0800, konsolebox wrote: > On Wed, Jul 13, 2016 at 1:30 AM, Philip Webb wrote: > > It looks as if in the near future I am going to have to install KDE 5 , > > if I want to go on using my regular daily apps Konsole Gwenview Okular ; > > yes, I know I can limit exposure to their requirements > > & don't need to install the whole desktop system. > > > > Before I plunge into that, is anyone else using KDE 5 every day ? > > What are people's experiences with it ? > > I just installed KDE 5 in order to try how one application works on > it, and also due to my curiosity. The applications that depend on it: > konsole and dolphin, doesn't work well if KDE5 is not itself the one > that's running. In dolphin, some icons don't show. konsole also > doesn't show some icons when it uses them as its own icon. After 4.14 packages left portage, I did bite the bullet (after rsyncing / to an external HDD). I’m not very fond of the new minimalist design language everyone seems to adopt, so I kept oxygen icons. I had to uninstall some blockers by hand (bye bye knemo, crystal and qtcurve[windeco]. :'-( ). At first I wanted to write a rant mail about lots of bugs after I made the world upgrade, but then I noticed that I only upgraded what was there, and did not install plasma-meta to pull in the new stuff. Once that was done, things looked a lot better. No hangups so far, not even a proper crash. I only disabled file indexing from the start because baloo was hogging my spinning drive. I might even try and find out what of the old akonadi indexing database I could delete and free up oogles of megabytes. I imagine that’s probably less hassle than start with a clean user directory and add all my PIM stuff back (several accounts for mail, contacts and calendar with all their detailed settings). What I do like: - the circular system monitor - actually less waste of space in the panel in comparison to Oxygen due to smaller margins around taskbar items and the pager widget What I miss: - no more eyes for the panel - no way of navigating the classical Kmenu via shortcut keys due to the always-on search field. What is the purpose of krunner then. - no good network monitor (yet?). The KDE-own has no options at all, I can't set anything (things like width, animation step size or hiding the interface name, which overlaps with the graph label). > Firefox seems to have been affected by it as well. It doesn't show the > scroll bar button, and some widget borders looked a little different. That's GTK3. I installed clearlooks-phenix to get back a more traditional look and also disabled their stupid non-standard default that -- when you left-click a scrollbar -- it scrolls directly to that position. After years of strong dislike of anything GTK due to ugliness in the mid-2000s, I came to terms with it a few years back, I even like XFCE on low-power devices. But GTK3 gets me all going again. Whenever I save a file from within Firefox, I am appalled by the file dialog and its non-existing usability. -- Gruß | Greetings | Qapla’ Please do not share anything from, with or about me with any social network. Latin: the late revenge of the Romans to all Germans. signature.asc Description: Digital signature
[gentoo-user] Re: KDE 5
On 13/07/16 07:56, Mick wrote: > Unlike Neil I found Plasma 5 a major climb down from KDE4 in terms of > interface usability. A lot of things were broken and for me still are. For > example Konqueror on my laptop does not integrate with Dolphin anymore, menu > icons on Dolphin are not shown, Network places are empty, Kim4 does not show > up in the dropdown menu, etc. I noticed the same problems on a desktop which > had the full Plasma5 desktop installed. These are not related to Plasma, but rather KDE Applications which is now distinct. The Dolphin/Konqueror integration should be fixed in 16.04.3, see bug #562224. For missing icons and network places, it would be nice to know more. It is expected that kim4 (kdelibs4-based) does not work with KF5-based Dolphin.
[gentoo-user] Portage getting slicker?
>
So, today I ran a sync and upgrade to a gentoo workstation::
emerge -uvDNp world
These are the packages that would be merged, in order:
Calculating dependencies... done!
Total: 0 packages, Size of downloads: 0 KiB
WARNING: One or more updates/rebuilds have been skipped due to a dependency
conflict:
media-libs/jasper:0
(media-libs/jasper-1.900.1-r9:0/0::gentoo, ebuild scheduled for merge)
conflicts with
media-libs/jasper:0/0=[abi_x86_32(-),abi_x86_64(-)] required by
(x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed)
media-libs/jasper:=[abi_x86_32(-),abi_x86_64(-)] required by
(x11-libs/gdk-pixbuf-2.32.3:2/2::gentoo, installed)
OK, no big deal, except I next ran::
'emerge -uDtv @world'
BANG, it just started compiling as if no blokers?
Calculating dependencies... done!
[ebuild rR] media-libs/mesa-11.0.6::gentoo USE="classic dri3 egl
gallium gbm gles2 llvm nptl openmax udev vaapi vdpau xvmc -bindist -d3d9
-debug -gles1 -opencl -osmesa -pax_kernel -pic (-selinux) -wayland -xa"
ABI_X86="32 (64) (-x32)" VIDEO_CARDS="radeon radeonsi (-freedreno) -i915
-i965 -ilo -intel -nouveau -r100 -r200 -r300 -r600 -vmware" 7,103 KiB
[nomerge ] sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo
[3.5.0:0/3.5::gentoo] USE="clang libffi ncurses python static-analyzer
-debug -doc -gold -libedit -lldb% -multitarget -ocaml {-test} -xml"
ABI_X86="32 (64) (-x32)" PYTHON_TARGETS="python2_7 (-pypy%)"
VIDEO_CARDS="radeon"
[blocks b ] <=sys-devel/clang-3.7.1-r99 ("<=sys-devel/clang-3.7.1-r99"
is blocking sys-devel/llvm-3.7.1-r3)
[ebuild U ] sys-devel/clang-3.7.1-r100:0/3.7::gentoo
[3.5.0-r100:0/3.5::gentoo] USE="python static-analyzer -debug -multitarget"
ABI_X86="(64) -32 (-x32)" 0 KiB
[ebuild r U ]sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo
[3.5.0:0/3.5::gentoo] USE="clang libffi ncurses python static-analyzer
-debug -doc -gold -libedit -lldb% -multitarget -ocaml {-test} -xml"
ABI_X86="32 (64) (-x32)" PYTHON_TARGETS="python2_7 (-pypy%)"
VIDEO_CARDS="radeon" 24,610 KiB
Total: 3 packages (2 upgrades, 1 reinstall), Size of downloads: 31,713 KiB
Conflict: 1 block
The following packages are causing rebuilds:
(sys-devel/llvm-3.7.1-r3:0/3.7.1::gentoo, ebuild scheduled for merge)
causes rebuilds for:
(media-libs/mesa-11.0.6:0/0::gentoo, ebuild scheduled for merge)
>>> Verifying ebuild manifests
>>> Running pre-merge checks for sys-devel/llvm-3.7.1-r3
* Checking for at least 2400 MiB disk space at
"/var/tmp/portage/sys-devel/llvm-3.7.1-r3/temp" ...
[ ok ]
>>> Emerging (1 of 3) sys-devel/llvm-3.7.1-r3::gentoo
Now this is col
I did nothing manual in between. Explanations?
James
[gentoo-user] Re: Gitlab experiences
"J." García gmail.com> writes: > > I know the Gentoo Infra team has had negative experiences with > > hosting > > just about anything Java and don't want to go near it. I don't know > > if that is based on specific experiences with GitLab or with just > > avoidance with Java in general. Most of the competing solutions in > > this space are also Java-based which is why we don't host any kind of > > alternative to Github. > What java has to do with gitlab? according to the repo I see is mostly > ruby code[1](both gitlab and gitlab-ci). what you wrote make it seems > like it is a java app. > In the github mirror of gitlab(the main app), the file stats are: > 2,253 Ruby > 697 Haml > 319 Markdown > 158 CoffeeScript > 99 SCSS > 90 Cucumber > 40 YAML > 39 HTML+ERB > 26 SVG > 25 JavaScript Huh. (Double huh...) As Alan is always quick to quip:: JavaScript ain't java, so that changes the entire game. Performance can be fixed with a gentoo cluster (yet to be proven). I also read that gitlab is working to make the "engine" faster. > I have tried it using docker, and it does uses a lot of resources as > someone previously said. That's what clusters are for. ymmv. Lots of folks are using mesos, for a variety of workloads. Others like Openstack, which is supported in gentoo portage. My work on Apache-Mesos is suspended until I finish up a rack install system I've been hacking on for a while. Apache-Mesos, that works, can be found in BGO. Folks are building mesos cluster on Arm64v8 (Rpi-3) and soon those arm clusters are going to everywhere, particularly in small companies An 8 node cluster for less than $500 USD. (includes 32 cores, 8 arm gpus and 16 gig of ram. So I'm just waiting on an arm64v8 board that has 4 gig of ram each. > BTW, after I quick look at Gogs it does seem a nice option, very > 'github like'. I will try it for sure. For me, robust gentoo support for both Gogs and Gitlab is a good idea as these are excellent codes to put on a (gentoo) cluster > [1] https://github.com/gitlabhq/gitlabhq hth, James
Re: [gentoo-user] Gitlab experiences
On Wed, Jul 13, 2016 at 03:19:00PM +0200, Ralf wrote: > Hi James, > > On 07/13/16 14:44, James wrote: > > > > Has anyone attempted to install a self hosted gitlab on gentoo server(s)? > > A small gentoo cluster/container setup? Using a Distributed File System, > > like cephfs, orangefs or other DFS? > I recommend to deploy gitlab inside a Debian LXC/Docker container as > Gitlab guys provide and maintain precompiled .deb packages. You do not > want to compile it on your own as it comes with a load of dependencies. > And once dependencies change you really might run into trouble with > gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's > a whole ecosystem. I would deploy it with docker. The gitlab guys push official images of the main gitlab app[1] and CI runners[2] to dockerhub. That should be the easiest path to getting it up and running in no time. That being said, gitlab does not really play well with clustering in general. I don't think the main part of the app does any kind of horizontal scaling (gitlab.com is hosted on a single server) so you need a fairly beefy server. And while storage should be entirely up to you (the app _should_ be indifferent to what you use) most folks appear to run with local disks or NFS. > > > > Any experiences with gitlab are most welcome for comment, good or bad. > Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with > large repos (except you have unlimited access to hardware resources). A > Linux kernel git mirror finally crashed it. I can second that - it's slow and load times are long even when performing the most basic operations such as opening a small file or viewing a single commit (you _will_ notice - it regulary takes several seconds). Mind you the gitlab folks are working on improving this since several releases. As a user, another issue I have with it is that the merge request/review interface is just terrible. There is _no_ merge request versioning, so either you submit your code in perfect shape at first try, or any change (amending/rebasing/merging) will cause the changelist to be duplicated many times over. You also lose track of the review history instantly - old comments are either concealed or swallowed. There is also no CLI utility to automate common review-related tasks (submitting/responding to review) so you are forced to do everything over the slow WebUI. If you care at all about the codereview aspects, I would recommend gerrit or phabricator. Both have cli utilities (git-review and arcanist) and while some claim they are ugly (heard that one especially about gerrit) they are 100x more practical. If you only care about having a repository browser then gitlab can work but there are simpler apps out there (gogs/pagure). [1]: https://hub.docker.com/r/gitlab/gitlab-ce/ [2]: https://hub.docker.com/r/gitlab/gitlab-runner/
Re: [gentoo-user] no sound through headphones (OK through speakers)
On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb wrote: > Hardware: Dell Latitude E7450 laptop > > Gentoo essentially all stable > Gnome / Systemd > > When I play a movie using totem it sounds fine if no headphones are > plugged it. > > No sound at all with headphones (I tried three different ones). > > The sound settings gui recognizes that headphones are in. The sound > test is silent. If I select the internal speakers in the gui (with the > headphones still in) sound is fine. > > Is there some headphone option I must enable in the kernel or elsewhere? > > thanks in advance. > allan > > PS I also tried cat /boot/vmlinux > /dev/audio1. Silent on headphones It will remember the volume/mute settings per device. Is it unmuted and turnt up?
Re: [gentoo-user] no sound through headphones (OK through speakers)
On Wed, Jul 13, 2016 at 12:55 PM, allan gottlieb wrote: > Hardware: Dell Latitude E7450 laptop > > Gentoo essentially all stable > Gnome / Systemd > > When I play a movie using totem it sounds fine if no headphones are > plugged it. > > No sound at all with headphones (I tried three different ones). > > The sound settings gui recognizes that headphones are in. The sound > test is silent. If I select the internal speakers in the gui (with the > headphones still in) sound is fine. > > Is there some headphone option I must enable in the kernel or elsewhere? The volume is probably muted for the headphones. Install pavucontrol, and execute it while the movie is playing. In the "Output Devices" tab look for your sound card (probably something like "Built-in Audio"), and in port select "Headphones". Then adjust the volume. Regards. -- Dr. Canek Peláez Valdés Profesor de Carrera Asociado C Departamento de Matemáticas Facultad de Ciencias Universidad Nacional Autónoma de México
[gentoo-user] no sound through headphones (OK through speakers)
Hardware: Dell Latitude E7450 laptop Gentoo essentially all stable Gnome / Systemd When I play a movie using totem it sounds fine if no headphones are plugged it. No sound at all with headphones (I tried three different ones). The sound settings gui recognizes that headphones are in. The sound test is silent. If I select the internal speakers in the gui (with the headphones still in) sound is fine. Is there some headphone option I must enable in the kernel or elsewhere? thanks in advance. allan PS I also tried cat /boot/vmlinux > /dev/audio1. Silent on headphones
Re: [gentoo-user] executing a command as a nologin user
Fernando Rodriguez wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA256 > > On 07/13/2016 07:10 AM, Alan McKinnon wrote: > > On 12/07/2016 03:47, jens w wrote: > >> .procmailrc > >> :0 c > >> * !^X-Loop: n...@example.com > >> | formail -X "From:" | $HOME/bin/script.sh > >> > >> procmail.log > >> procmail: Executing " formail -X "From:" | $HOME/bin/script.sh > >> > >> for incoming mail, a script is executed. logfile has the same > >> entry as it is in other users. but the script do nothing. > >> > >> How executing a command as a nologin user? > >> > > > > > > You can't, not the way you are doing it. > > You want to launch a shell script for the user, but the user's > > shell is /sbin/nologin. This exits immediately without launching > > the script. > > > > Give the user a real shell. > > > > Alan > > > > I've been following this thread and thinking the same thing but > wasn't sure. > > What if you invoke the shell directly instead of the script, either: > /bin/sh -c "" or /bin/sh -c "$(cat
Re: [gentoo-user] KDE 5
On Wednesday 13 Jul 2016 09:14:37 Stefano Crocco wrote: > On Tuesday 12 July 2016 22:56:38 Mick wrote: > > For example Konqueror on my laptop does not integrate with Dolphin > > anymore, > > This is because konqueror still uses KDE 4, while dolphin has been ported to > KDE Frameworks 5. The problem is that konqueror has been without a full > time mantainer for a long time, since the original author shifted his > attention to other parts of KDE. Since last year I took up the role of > konqueror mantainer, but unfortunately I had much less free time to work on > it than I hoped for back then. > > Currently, the situation regarding konqueror is the following: > > * the application itself has been ported from KDE4/Qt4 to KDE Frameworks/Qt5 > by its original author (David Faure). This means that it compiles and > mostly works, including the integration with dolphin. There are several > bugs, however. > * I've ported most of the plugins (only the user agent plugin doesn't work > completely) > * I'm in the process of porting the sidebar, which is the last part of > konqueror still using KDE 4. However, different from the rest of the > application, porting this is going to be a lot of work. This is because the > sidebar (or rather, parts of it) still uses a compatibility library used to > port applications from KDE 3 to KDE 4. This library, of course, doesn't > exist in KDE Frameworks, which means that a large part of the sidebar code > needs to be almost rewritten. > > I've thought of writing an ebuild for the developer version of konqueror, (a > one, of course), since the application is currently mostly working (I > use it as my primary browser, even if I have to switch to other browser > every now and then for sites which won't work with it). The reason I > haven't done so is that konqueror requires libkonq and that the 4 and 5 > versions of libkonq can't be installed together. Since many applications > from kde 4 still use it, this would create an issue. > > Stefano Thank you Stefano for taking this project on! We are grateful for your work. I have been using Konqueror since KDE2 if I recall right, and although I've tried other browsers/file managers/ftp clients/et al. nothing has ever come close in terms of versatility to this 'swiss army knife' on all my Linux desktops. From what you say it seems that libkonq-5 won't become available until all reverse deps of KDE4 applications are no longer there. If this can't be addressed by slots in Gentoo, I guess we'll just have to wait until all KDE4 applications have been ported to KDE5, or retired. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Using SSH around the LAN
On Wednesday 13 Jul 2016 09:48:59 Peter Humphrey wrote: > On Tuesday 12 July 2016 17:48:33 Alan McKinnon wrote: > > On 12/07/2016 17:42, Peter Humphrey wrote: > > > Is there a guide to setting up password-less authentication to enable me > > > to do this? > > > > http://www.funtoo.org/Keychain > > Thanks Alan. I don't think it's the one I read before but it looks useful > anyway. > > > Note that you, portage and root are 3 different users, so you must make > > key pairs for each on each source machine you will ssh from. > > > > Then you need to add each of those user's public keys to each > > destination user's authorized_keys file on each machine you want to ssh > > to. > > > > That can be a lot of key copying :-) 3 x 3 x # of machines > > > > Finally, on each machine you will ssh from and as each user who will do > > the ssh'ing, you must run keychain at least once to store the key creds. > > They should then persist until reboot, when you must run keychain again > > for each user. > > Hmm. I may end up just allowing ssh password authentication and relying on > my vDSL router to keep other people's noses out of my business. The portage > user can't log in anyway, so its scp-ing and rsyncing would have to be done > by root. > > > The idea is that a given user's keychain creds are valid over all that > > user's login sessions on a machine. Users cannot share each other's > > keychain > > You've given me plenty to think about - thanks again. Something else to think about is to only allow the login shell to execute limited command(s), for example to only be able to su to portage and run rsync or some such. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] executing a command as a nologin user
Alan McKinnon wrote: > On 13/07/2016 18:42, wabe wrote: > > Fernando Rodriguez wrote: > > > >> -BEGIN PGP SIGNED MESSAGE- > >> Hash: SHA256 > >> > >> On 07/13/2016 07:10 AM, Alan McKinnon wrote: > >>> On 12/07/2016 03:47, jens w wrote: > .procmailrc > :0 c > * !^X-Loop: n...@example.com > | formail -X "From:" | $HOME/bin/script.sh > > procmail.log > procmail: Executing " formail -X "From:" | $HOME/bin/script.sh > > for incoming mail, a script is executed. logfile has the same > entry as it is in other users. but the script do nothing. > > How executing a command as a nologin user? > > >>> > >>> > >>> You can't, not the way you are doing it. > >>> You want to launch a shell script for the user, but the user's > >>> shell is /sbin/nologin. This exits immediately without launching > >>> the script. > >>> > >>> Give the user a real shell. > >>> > >>> Alan > >>> > >> > >> I've been following this thread and thinking the same thing but > >> wasn't sure. > > > > I don't think so. To proof it, I created this user: > > > > nologinuser:x:1015:1016::/home/nologinuser:/sbin/nologin > > > > Then I created this script: > > > > #!/bin/sh > > # > > date >> /home/nologinuser/envars.txt > > echo $HOME >> /home/nologinuser/envars.txt > > echo $PATH >> /home/nologinuser/envars.txt > > echo "---" >> /home/nologinuser/envars.txt > > > > I stored it as /var/script-nologinuser/testscript.sh. I had to store > > it at this place because /home is mounted with the noexec option on > > my system. > > > > > > Then I created a cronjob for the nologinuser user: > > > > */2 * * * * /var/script-nologinuser/testscript.sh > > > > > > The result is the file /home/nologinuser/envars.txt with this > > content: > > > > Wed Jul 13 18:10:01 CEST 2016 > > /home/nologinuser > > /usr/bin:/bin > > --- > > Wed Jul 13 18:12:01 CEST 2016 > > /home/nologinuser > > /usr/bin:/bin > > --- > > > > > > The ownership and the rights for /home/nologinuser/, > > /var/script-nologinuser/ and /var/script-nologinuser/testscript.sh > > are nologinuser:nologinuser and 700. > > > > So it seems, that it is possible to execute scripts without setting > > a shell in /etc/passwd. > > > > I don't know why it doesn't work for jens w. > > > Not so. Your script is launched by cron, running as root. It starts a > non-interactive no-login shell (that's why people have infernal > trouble with cron, assuming it has a $PATH when it actually has none) Ok, this seems to make sense. I did another test to make sure that it works also when the script isn't executed directly by crontab. I'm not sure if this is a real proof, because the "execution chain" is started by cron. I renamed /var/script-nologinuser/testscript.sh to /var/script-nologinuser/testscript-2.sh and created a script /var/script-nologinuser/testscript.sh with this content: #!/bin/sh # /var/script-nologinuser/testscript-2.sh It also worked. -- Regards wabe
Re: [gentoo-user] executing a command as a nologin user
On 13/07/2016 18:42, wabe wrote: Fernando Rodriguez wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 07/13/2016 07:10 AM, Alan McKinnon wrote: On 12/07/2016 03:47, jens w wrote: .procmailrc :0 c * !^X-Loop: n...@example.com | formail -X "From:" | $HOME/bin/script.sh procmail.log procmail: Executing " formail -X "From:" | $HOME/bin/script.sh for incoming mail, a script is executed. logfile has the same entry as it is in other users. but the script do nothing. How executing a command as a nologin user? You can't, not the way you are doing it. You want to launch a shell script for the user, but the user's shell is /sbin/nologin. This exits immediately without launching the script. Give the user a real shell. Alan I've been following this thread and thinking the same thing but wasn't sure. I don't think so. To proof it, I created this user: nologinuser:x:1015:1016::/home/nologinuser:/sbin/nologin Then I created this script: #!/bin/sh # date >> /home/nologinuser/envars.txt echo $HOME >> /home/nologinuser/envars.txt echo $PATH >> /home/nologinuser/envars.txt echo "---" >> /home/nologinuser/envars.txt I stored it as /var/script-nologinuser/testscript.sh. I had to store it at this place because /home is mounted with the noexec option on my system. Then I created a cronjob for the nologinuser user: */2 * * * * /var/script-nologinuser/testscript.sh The result is the file /home/nologinuser/envars.txt with this content: Wed Jul 13 18:10:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- Wed Jul 13 18:12:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- The ownership and the rights for /home/nologinuser/, /var/script-nologinuser/ and /var/script-nologinuser/testscript.sh are nologinuser:nologinuser and 700. So it seems, that it is possible to execute scripts without setting a shell in /etc/passwd. I don't know why it doesn't work for jens w. Not so. Your script is launched by cron, running as root. It starts a non-interactive no-login shell (that's why people have infernal trouble with cron, assuming it has a $PATH when it actually has none)
Re: [gentoo-user] executing a command as a nologin user
Fernando Rodriguez wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA256 > > On 07/13/2016 07:10 AM, Alan McKinnon wrote: > > On 12/07/2016 03:47, jens w wrote: > >> .procmailrc > >> :0 c > >> * !^X-Loop: n...@example.com > >> | formail -X "From:" | $HOME/bin/script.sh > >> > >> procmail.log > >> procmail: Executing " formail -X "From:" | $HOME/bin/script.sh > >> > >> for incoming mail, a script is executed. logfile has the same > >> entry as it is in other users. but the script do nothing. > >> > >> How executing a command as a nologin user? > >> > > > > > > You can't, not the way you are doing it. > > You want to launch a shell script for the user, but the user's > > shell is /sbin/nologin. This exits immediately without launching > > the script. > > > > Give the user a real shell. > > > > Alan > > > > I've been following this thread and thinking the same thing but > wasn't sure. I don't think so. To proof it, I created this user: nologinuser:x:1015:1016::/home/nologinuser:/sbin/nologin Then I created this script: #!/bin/sh # date >> /home/nologinuser/envars.txt echo $HOME >> /home/nologinuser/envars.txt echo $PATH >> /home/nologinuser/envars.txt echo "---" >> /home/nologinuser/envars.txt I stored it as /var/script-nologinuser/testscript.sh. I had to store it at this place because /home is mounted with the noexec option on my system. Then I created a cronjob for the nologinuser user: */2 * * * * /var/script-nologinuser/testscript.sh The result is the file /home/nologinuser/envars.txt with this content: Wed Jul 13 18:10:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- Wed Jul 13 18:12:01 CEST 2016 /home/nologinuser /usr/bin:/bin --- The ownership and the rights for /home/nologinuser/, /var/script-nologinuser/ and /var/script-nologinuser/testscript.sh are nologinuser:nologinuser and 700. So it seems, that it is possible to execute scripts without setting a shell in /etc/passwd. I don't know why it doesn't work for jens w. -- Regards wabe
Re: [gentoo-user] Re: Konsole
On 07/12/2016 03:29 PM, konsolebox wrote: > 1) kde-base/kactivitymanagerd-4.13.3-r1 was just removed and not even > copied to kde-sunset. > 2) kde-base/kactivities-4.13.3-r2 was just modified with a dependency > to kde-plasma/kactivitymanagerd:5. Not even a revision bump was made. > It was also not copied to kde-sunset. > I sync'ed today and found this: >From https://anongit.gentoo.org/git/proj/kde-sunset 30a5f0c..302b10b master -> origin/master Updating 30a5f0c..302b10b Fast-forward kde-base/kactivities/Manifest | 1 + kde-base/kactivities/kactivities-4.13.3-r2.ebuild | 26 ++ kde-base/kactivities/metadata.xml | 5 + kde-base/kactivitymanagerd/Manifest | 1 + kde-base/kactivitymanagerd/kactivitymanagerd-4.13.3-r1.ebuild | 31 +++ kde-base/kactivitymanagerd/metadata.xml | 5 + 6 files changed, 69 insertions(+) create mode 100644 kde-base/kactivities/Manifest create mode 100644 kde-base/kactivities/kactivities-4.13.3-r2.ebuild create mode 100644 kde-base/kactivities/metadata.xml create mode 100644 kde-base/kactivitymanagerd/Manifest create mode 100644 kde-base/kactivitymanagerd/kactivitymanagerd-4.13.3-r1.ebuild create mode 100644 kde-base/kactivitymanagerd/metadata.xml So they copied the offending packages to sunset. I've removed kactivities from the local overlay, and: # emerge -pvuDN world These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild R] kde-base/kactivities-4.13.3-r2:4/4.13::kde-sunset [4.13.3-r2:4/4.13::local] USE="(-aqua)" 0 KiB Total: 1 package (1 reinstall), Size of downloads: 0 KiB So it works again, just FYI for others. Dan
Re: [gentoo-user] Re: Gitlab experiences
Hi, On 07/13/16 17:30, James wrote: > Ralf ramses-pyramidenbau.de> writes: > > >>> From the gentoo wiki, it looks like all of the dependencies are already >>> in portage:: >>> https://wiki.gentoo.org/wiki/GitLab#Prerequisites.2FDependencies Even if all dependency stuff is already inside portage, gitlab itself is not. So there's no guarantee that gitlab will meet dependency requirements in future when some packages upgrade or APIs change. > > >> I recommend to deploy gitlab inside a Debian LXC/Docker container as >> Gitlab guys provide and maintain precompiled .deb packages. You do not >> want to compile it on your own as it comes with a load of dependencies. >> And once dependencies change you really might run into trouble with >> gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's >> a whole ecosystem. > > I wonder if using 'app-arch/dpkg' to just install the .deb files on gentoo > would work? > Issues with using dpkg to install gitlab on gentoo inside a VM/container ? Interesting idea, but I never tried it. The gitlab .deb is pretty huge (~350MiB) and is shipped along with its own webserver, ruby and all that other dependency monsters and a minimum requirement to external dependencies. So maybe this is in deed worth a try! Good point. But what I know for sure that the .deb said to fit to current debian, where i have kind of a 'guarantee' that it will work in future with minimum pain. So for me a debian LXC container inside my gentoo box is the best solution. > > >> For private use, I deployed my Gitlab inside a LXC container on my >> Gentoo server box, everything else is really way too much tinkering >> around. If you have no other problems in your life, just try it out and >> go for it. > > Can you be more specific? Exactly which package(s) did you install this way? > Is debian inside that LXC, or just pure gentoo? Are you using CI with this > configuration? Yes, it's debian inside LXC. And no, I'm not using CI. > > >>> Any experiences with gitlab are most welcome for comment, good or bad. >> Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with >> large repos (except you have unlimited access to hardware resources). A >> Linux kernel git mirror finally crashed it. > > On a cluster would be my approach, after the installation issues are ironed > out on a single server install. > >> That's why I decided to switch to Gogs [1], even for business cases. >> Gogs is implemented in Go, has a pretty active and responsive community >> and (in my opinion) it is a well-maintained project. Looks and feels >> like gitlab but is much faster and consumes a minimum of resources. I >> strongly recommend to use Gogs. Just try it out on their website. > > Is this the gogs package you installed:: www-apps/gogs [1] > [1] "go-overlay" layman/go-overlay > > Or did you just use a SaaS/PaaS for Gogs? Same here (sorry i was inaccurate): Gogs guys also provide a .deb file. Same strategy: Debian inside a LXC container on my Gentoo box. > > >> They also provide a .deb package, that's the reason why I'm running it >> inside a Debian LXC container as well. > > And this runs on a gentoo server, with debian inside the LXC? Or on a debian > machine with LXC? Ok, so this is my *private* setup: Single server box with gentoo on bare-metal, latest bleeding edge stuff. Nginx on that Gentoo that serves some lightweight sites and webapps (wordpress, roundcube, usual suspects). Nginx also terminates SSL. I did not want to install gitlab on gentoo because of its tons of dependencies. So I run a Debian LXC Container inside Gentoo. Nginx then reverse-proxies and SSL-terminates the Gitlab LXC container and iptables forwards the ssh port of the gitlab container. Updating gitlab to the latest version just costs an apt-get upgrade and a dozen updates later nothing exploded so far. > > Have you tried any VCS on a cluster (openstack/mesos/hadoop/others? Nope, not my department. Ralf > > curiously, > James > > > > >
[gentoo-user] Re: Gitlab experiences
Ralf ramses-pyramidenbau.de> writes: > > From the gentoo wiki, it looks like all of the dependencies are already > > in portage:: > > https://wiki.gentoo.org/wiki/GitLab#Prerequisites.2FDependencies > I recommend to deploy gitlab inside a Debian LXC/Docker container as > Gitlab guys provide and maintain precompiled .deb packages. You do not > want to compile it on your own as it comes with a load of dependencies. > And once dependencies change you really might run into trouble with > gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's > a whole ecosystem. I wonder if using 'app-arch/dpkg' to just install the .deb files on gentoo would work? Issues with using dpkg to install gitlab on gentoo inside a VM/container ? > For private use, I deployed my Gitlab inside a LXC container on my > Gentoo server box, everything else is really way too much tinkering > around. If you have no other problems in your life, just try it out and > go for it. Can you be more specific? Exactly which package(s) did you install this way? Is debian inside that LXC, or just pure gentoo? Are you using CI with this configuration? > > Any experiences with gitlab are most welcome for comment, good or bad. > Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with > large repos (except you have unlimited access to hardware resources). A > Linux kernel git mirror finally crashed it. On a cluster would be my approach, after the installation issues are ironed out on a single server install. > That's why I decided to switch to Gogs [1], even for business cases. > Gogs is implemented in Go, has a pretty active and responsive community > and (in my opinion) it is a well-maintained project. Looks and feels > like gitlab but is much faster and consumes a minimum of resources. I > strongly recommend to use Gogs. Just try it out on their website. Is this the gogs package you installed:: www-apps/gogs [1] [1] "go-overlay" layman/go-overlay Or did you just use a SaaS/PaaS for Gogs? > They also provide a .deb package, that's the reason why I'm running it > inside a Debian LXC container as well. And this runs on a gentoo server, with debian inside the LXC? Or on a debian machine with LXC? Have you tried any VCS on a cluster (openstack/mesos/hadoop/others? curiously, James
Re: [gentoo-user] Gitlab experiences
El mié, 13-07-2016 a las 09:13 -0400, Rich Freeman escribió: > On Wed, Jul 13, 2016 at 8:44 AM, James > wrote: > > > > Has anyone attempted to install a self hosted gitlab on gentoo > > server(s)? > > A small gentoo cluster/container setup? Using a Distributed File > > System, > > like cephfs, orangefs or other DFS? > > > > I know the Gentoo Infra team has had negative experiences with > hosting > just about anything Java and don't want to go near it. I don't know > if that is based on specific experiences with GitLab or with just > avoidance with Java in general. Most of the competing solutions in > this space are also Java-based which is why we don't host any kind of > alternative to Github. > What java has to do with gitlab? according to the repo I see is mostly ruby code[1](both gitlab and gitlab-ci). what you wrote make it seems like it is a java app. In the github mirror of gitlab(the main app), the file stats are: 2,253 Ruby 697 Haml 319 Markdown 158 CoffeeScript 99 SCSS 90 Cucumber 40 YAML 39 HTML+ERB 26 SVG 25 JavaScript I have tried it using docker, and it does uses a lot of resources as someone previously said. BTW, after I quick look at Gogs it does seem a nice option, very 'github like'. I will try it for sure. [1] https://github.com/gitlabhq/gitlabhq
[gentoo-user] Re: Gitlab experiences
Ralf ramses-pyramidenbau.de> writes: > On 07/13/16 16:35, Rich Freeman wrote: > > On Wed, Jul 13, 2016 at 9:19 AM, Ralf > > ramses-pyramidenbau.de> wrote: > >> I recommend to deploy gitlab inside a Debian LXC/Docker container as > >> Gitlab guys provide and maintain precompiled .deb packages. You do not > >> want to compile it on your own as it comes with a load of dependencies. > >> And once dependencies change you really might run into trouble with > >> gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's > >> a whole ecosystem. Yep, I heard a bit of bitching about maintaining it. I was contacted by a corp to install and manage gitlab for them. It's not my cup of tea, so folks in that space must be upset with maintenance issues > > This is part of why I'd suggest not upgrading it in-place. Just > > create a new container from scratch every time there is an update. > Sure, I totally agree. But from a maintenance point of view this can > become a full-time job very quickly, as gitlab has pretty short > innovation and release cycles. And you do want to install updates very > quickly, especially for web-apps. Interesting idea. It certainly is a good candidate for codes that stress a cluster. I'll keep that in mind. > Ralf > > OF course, Gentoo makes this somewhat more painful than other distros; > > it tends to be designed around upgrading in-place. I did not realize it was so java centric.. I'm out, cause I have more icedtea-java projects than I know what to do with. (apache-spark). Thanks for all the info guys, James
Re: [gentoo-user] Gitlab experiences
Hi Rich, On 07/13/16 16:35, Rich Freeman wrote: > On Wed, Jul 13, 2016 at 9:19 AM, Ralf > wrote: >> I recommend to deploy gitlab inside a Debian LXC/Docker container as >> Gitlab guys provide and maintain precompiled .deb packages. You do not >> want to compile it on your own as it comes with a load of dependencies. >> And once dependencies change you really might run into trouble with >> gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's >> a whole ecosystem. > > This is part of why I'd suggest not upgrading it in-place. Just > create a new container from scratch every time there is an update. Of Sure, I totally agree. But from a maintenance point of view this can become a full-time job very quickly, as gitlab has pretty short innovation and release cycles. And you do want to install updates very quickly, especially for web-apps. Ralf > course, Gentoo makes this somewhat more painful than other distros; it > tends to be designed around upgrading in-place.
Re: [gentoo-user] Gitlab experiences
On Wed, Jul 13, 2016 at 9:19 AM, Ralf wrote: > I recommend to deploy gitlab inside a Debian LXC/Docker container as > Gitlab guys provide and maintain precompiled .deb packages. You do not > want to compile it on your own as it comes with a load of dependencies. > And once dependencies change you really might run into trouble with > gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's > a whole ecosystem. This is part of why I'd suggest not upgrading it in-place. Just create a new container from scratch every time there is an update. Of course, Gentoo makes this somewhat more painful than other distros; it tends to be designed around upgrading in-place. -- Rich
Re: [gentoo-user] Gitlab experiences
Hi James, On 07/13/16 14:44, James wrote: > Hello, > > So, from what I've read, gitlab is very popular because you can self-host > and the CI capabilities there are built in. As I look at migrating several > several old mechanical ways to manage code inhouse, gitlab does look > appealing. I do see a variety of packages in via remote overlays:: > > [1] "cvut" layman/cvut > [2] "dev-zero" layman/dev-zero > [3] "gitlab" layman/gitlab > [4] "maksbotan" layman/maksbotan > [5] "mrueg" layman/mrueg > [6] "rindeal" layman/rindeal > [7] "R_Overlay" layman/R_Overlay > [8] "spike" layman/spike > > and dev-go/go-gitlab-client in portage. > > From the gentoo wiki, it looks like all of the dependencies are already > in portage:: > https://wiki.gentoo.org/wiki/GitLab#Prerequisites.2FDependencies > > > Has anyone attempted to install a self hosted gitlab on gentoo server(s)? > A small gentoo cluster/container setup? Using a Distributed File System, > like cephfs, orangefs or other DFS? I recommend to deploy gitlab inside a Debian LXC/Docker container as Gitlab guys provide and maintain precompiled .deb packages. You do not want to compile it on your own as it comes with a load of dependencies. And once dependencies change you really might run into trouble with gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's a whole ecosystem. For private use, I deployed my Gitlab inside a LXC container on my Gentoo server box, everything else is really way too much tinkering around. If you have no other problems in your life, just try it out and go for it. :-) > > Any experiences with gitlab are most welcome for comment, good or bad. Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with large repos (except you have unlimited access to hardware resources). A Linux kernel git mirror finally crashed it. That's why I decided to switch to Gogs [1], even for business cases. Gogs is implemented in Go, has a pretty active and responsive community and (in my opinion) it is a well-maintained project. Looks and feels like gitlab but is much faster and consumes a minimum of resources. I strongly recommend to use Gogs. Just try it out on their website. They also provide a .deb package, that's the reason why I'm running it inside a Debian LXC container as well. HTH Ralf [1] https://gogs.io/ > > -- Ralf Ramsauer PGP: 0x8F10049B
Re: [gentoo-user] Gitlab experiences
On Wed, Jul 13, 2016 at 8:44 AM, James wrote: > > Has anyone attempted to install a self hosted gitlab on gentoo server(s)? > A small gentoo cluster/container setup? Using a Distributed File System, > like cephfs, orangefs or other DFS? > I know the Gentoo Infra team has had negative experiences with hosting just about anything Java and don't want to go near it. I don't know if that is based on specific experiences with GitLab or with just avoidance with Java in general. Most of the competing solutions in this space are also Java-based which is why we don't host any kind of alternative to Github. So, take that as a data point or not as you wish. Certainly interested to hear what others have found. One thing I might suggest if there are concerns with maintenance of Java is that you try to containerize it as much as possible. Put Gitlab and its dependencies in a container. Use config management to carefully track what need to be changed above the baseline to get it working. Put all the storage on mounts hosted outside the container and bind/network mount them into the container. Then if things get out of hand in updates/etc you can just build a new Gitlab container on each update, apply configuration and mounts, and then test it out. You could make everything Gitlab depends on essentially disposable that way. -- Rich
[gentoo-user] Gitlab experiences
Hello, So, from what I've read, gitlab is very popular because you can self-host and the CI capabilities there are built in. As I look at migrating several several old mechanical ways to manage code inhouse, gitlab does look appealing. I do see a variety of packages in via remote overlays:: [1] "cvut" layman/cvut [2] "dev-zero" layman/dev-zero [3] "gitlab" layman/gitlab [4] "maksbotan" layman/maksbotan [5] "mrueg" layman/mrueg [6] "rindeal" layman/rindeal [7] "R_Overlay" layman/R_Overlay [8] "spike" layman/spike and dev-go/go-gitlab-client in portage. >From the gentoo wiki, it looks like all of the dependencies are already in portage:: https://wiki.gentoo.org/wiki/GitLab#Prerequisites.2FDependencies Has anyone attempted to install a self hosted gitlab on gentoo server(s)? A small gentoo cluster/container setup? Using a Distributed File System, like cephfs, orangefs or other DFS? Any experiences with gitlab are most welcome for comment, good or bad.
Re: [gentoo-user] Acceptable character encodings for files in /etc/portage?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 07/13/2016 08:08 AM, Fernando Rodriguez wrote: > On 07/11/2016 12:11 AM, P Levine wrote: >> Glep 31 states that text files in the portage tree should use UTF-8 > encoding. I see nothing indicating any standard encoding for files in > the /etc/portage directory. Since everything I have seen in portage's > python source-code seems to imply Unicode as valid but I can see nothing > specific to UTF-8, I thought I'd ask. Is there any official Gentoo standard > documentation that defines the acceptable set of character encodings for files > in /etc/portage? > > I just glanced over Glep 31 and it does not say that. It is talking about > ChangeLogs and metadata.xml (because they contain developer names with unicode > characters). For patches and such it says to follow upstream suit. > > If you try to use non-ASCII characters on ebuilds repoman will whine so I > suppose only ascii is supported and likewise there's nothing that I know of on > /etc/portage that requires non ascii characters. > > Actually, it does says it, my bad. But repoman does complains if you use non-ASCII characters and it says not to print non-ASCII so I guess it's talking about comments. -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQIcBAEBCAAGBQJXhjCoAAoJEPbOFX/5UlwcFzMP/RUKDiN5Svo8HQAzVm0GxZKd Mf9tIpDQUiUl+ZudSEVcpB8YdM5mcFY/a2VMLTjanppi6NLdQlDywIOWET7e8uSs X4qR0D2bxDDs8TtM1aiWSa5EzeJ42eNIN/qLFAImWDluZFNxokXt78aaEX5O1o+u dYsygCnMzjRFc0y9JxoZpYWKq7tehHdBzxgeOHEzcu0m5DXxqoHWFxkqu1mr5Hjz lTe91gLsDTmqlttlqCGaac0NZMXZHh1SVtysJBlhryTOMWu/Xh1YQDP6SXX4Mnwi BvsSMXF4FrMvMaSxnn56DHDssR3fRFx595+DrU26GtNgTdBpmE24ymjsdcSm9Hm9 ovnG0h8Qys+pB15VUioNdR3LZTHMAO2/mx6lMYRdKvv735DEbIwCEXanSB9CG+nI VpOPaiYGunzDfsnV03ZRVFutbABM+cK70rBQeLUtH5FllDq9g3xEIMpdkradqRXC uXQTIoFslJxZH/AXGQTFCmIr6RGHdVfmmuHvp+NWGbwy3Q1DOfjcJK533lSoYScZ 8q2FDYdwidTDSCX1TaR5WKMpiwSdVftie9VOyotklk/MVukxHGmi8PH3guYqC3M5 ksy7AA0UrKXDiEaiVoMXNcP3FtJdWC6q+7a8PH59QOPDbZySP/HscGC7raFEMgNv JZ2NVoGuWXvwJMmJh2x9 =f29Q -END PGP SIGNATURE-
Re: [gentoo-user] Acceptable character encodings for files in /etc/portage?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 07/11/2016 12:11 AM, P Levine wrote: > Glep 31 states that text files in the portage tree should use UTF-8 encoding. I see nothing indicating any standard encoding for files in the /etc/portage directory. Since everything I have seen in portage's python source-code seems to imply Unicode as valid but I can see nothing specific to UTF-8, I thought I'd ask. Is there any official Gentoo standard documentation that defines the acceptable set of character encodings for files in /etc/portage? I just glanced over Glep 31 and it does not say that. It is talking about ChangeLogs and metadata.xml (because they contain developer names with unicode characters). For patches and such it says to follow upstream suit. If you try to use non-ASCII characters on ebuilds repoman will whine so I suppose only ascii is supported and likewise there's nothing that I know of on /etc/portage that requires non ascii characters. -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQIcBAEBCAAGBQJXhi9JAAoJEPbOFX/5UlwcDzEP/iea/YL+ZlyZ9Cg9FuknuxNF k/CZZFgUcHDfqFQswEsgARTfPEkP9TQkLqg53LswftcvfWcfs020HmOIRWe4h0uI gBprWHAH1cKb6sN8psEnYNFNgc7UBwguXfLvPOv4kEtsLsYVA+CQ9zj/Y82yGEMu yPtNXJUvQ1xyFMdWlXOwy0dt3DWcSozVLy8odjO8LP/LFH0c2lTtoY/UrtSFRt/l ASSbpj6m0A3KvtrQsCxYG1muht4YHePTRQ46vZdr1XbS8SIHhz8au+Ed9e5cK6CH 3+UNh4qKAbgugVxNjp7qnQxyth2yRMTJzyTw6LTk+p/5S2kF0gNmRUBYMgYgJ+KZ bmI1RsbjDeQXwngl8EA3wkBNOBl/6sq8QCNd7LxoCXGuNDRRDr2Eg2TIpOt90ZdQ 7rrr7G4LlrttByoLGzn85VFePbr8M52GmjN7la5kX+FFMwk+Ynx4HJSwbTHp6tMe q9ge1JznO6LHhGXzTzoKvY7xzvdEMGLItfSnZxvkD0xrU/hxpkiyoa7Xrmb35UR5 KmcQW7ZuXKYnWTjSCH9BUCNxkdd8z6WyGQUulVhNSn7Zyi7Edc/aP7lpkX5XBE3/ blif32JjinjcmlUSRfY6I+xYGDNfnChrJXq22uXz4laXxjIZ9qyUGsUVhC8s9rl3 W5KDa1rgHbNI7u0xnd/e =ayxA -END PGP SIGNATURE-
Re: [gentoo-user] executing a command as a nologin user
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 07/13/2016 07:10 AM, Alan McKinnon wrote: > On 12/07/2016 03:47, jens w wrote: >> .procmailrc >> :0 c >> * !^X-Loop: n...@example.com >> | formail -X "From:" | $HOME/bin/script.sh >> >> procmail.log >> procmail: Executing " formail -X "From:" | $HOME/bin/script.sh >> >> for incoming mail, a script is executed. logfile has the same entry as >> it is in other users. but the script do nothing. >> >> How executing a command as a nologin user? >> > > > You can't, not the way you are doing it. > You want to launch a shell script for the user, but the user's shell is > /sbin/nologin. This exits immediately without launching the script. > > Give the user a real shell. > > Alan > I've been following this thread and thinking the same thing but wasn't sure. What if you invoke the shell directly instead of the script, either: /bin/sh -c "" or /bin/sh -c "$(cat
Re: [gentoo-user] executing a command as a nologin user
On 12/07/2016 03:47, jens w wrote: .procmailrc :0 c * !^X-Loop: n...@example.com | formail -X "From:" | $HOME/bin/script.sh procmail.log procmail: Executing " formail -X "From:" | $HOME/bin/script.sh for incoming mail, a script is executed. logfile has the same entry as it is in other users. but the script do nothing. How executing a command as a nologin user? You can't, not the way you are doing it. You want to launch a shell script for the user, but the user's shell is /sbin/nologin. This exits immediately without launching the script. Give the user a real shell. Alan
Re: [gentoo-user] executing a command as a nologin user
Am Wed, 13 Jul 2016 03:02:50 +0200 schrieb wabe : > > Are you sure that the HOME envar is defined for the "nologin-user"? # getent passwd user user:x:1018:1018::'/home/user:/sbin/nologin # su -s /bin/bash user -c echo $HOME I change $HOME in /home/user, but the script not executed. > Is there a existing home directory defined in /etc/passwd for this > user? yes, and the mail-delivery to /home/user/.maildir ist ok, also the access from roundcube is ok.
Re: [gentoo-user] KDE 5
On Tuesday 12 July 2016 13:30:58 Philip Webb wrote: > It looks as if in the near future I am going to have to install KDE 5 , > if I want to go on using my regular daily apps Konsole Gwenview Okular ; > yes, I know I can limit exposure to their requirements > & don't need to install the whole desktop system. > > Before I plunge into that, is anyone else using KDE 5 every day ? > What are people's experiences with it ? I built a separate KDE-5 system in some spare disk space to see what it was like. I thought it was the ugliest desktop I'd ever seen, but after a good deal of poking around and waiting for enough development progress to have been made I did eventually find a group of settings that suited me. At that point, about a week ago, I converted my KDE-4 main system to KDE-5 and I'm now using it happily. This is a desktop box running BOINC projects 24/7 so it has no need of hibernation etc. but I do use KMail, which I've always liked. I also don't use dolphin much, nor konqueror at all, so I haven't stumbled over the problems some others are reporting. So I'm fairly happy with it for now. -- Rgds Peter
Re: [gentoo-user] Using SSH around the LAN
On Tuesday 12 July 2016 17:48:33 Alan McKinnon wrote: > On 12/07/2016 17:42, Peter Humphrey wrote: > > Is there a guide to setting up password-less authentication to enable me > > to do this? > > http://www.funtoo.org/Keychain Thanks Alan. I don't think it's the one I read before but it looks useful anyway. > Note that you, portage and root are 3 different users, so you must make > key pairs for each on each source machine you will ssh from. > > Then you need to add each of those user's public keys to each > destination user's authorized_keys file on each machine you want to ssh to. > > That can be a lot of key copying :-) 3 x 3 x # of machines > > Finally, on each machine you will ssh from and as each user who will do > the ssh'ing, you must run keychain at least once to store the key creds. > They should then persist until reboot, when you must run keychain again > for each user. Hmm. I may end up just allowing ssh password authentication and relying on my vDSL router to keep other people's noses out of my business. The portage user can't log in anyway, so its scp-ing and rsyncing would have to be done by root. > The idea is that a given user's keychain creds are valid over all that > user's login sessions on a machine. Users cannot share each other's > keychain You've given me plenty to think about - thanks again. -- Rgds Peter
Re: [gentoo-user] KDE 5
On Tuesday 12 July 2016 22:56:38 Mick wrote: > For example Konqueror on my laptop does not integrate with Dolphin > anymore, This is because konqueror still uses KDE 4, while dolphin has been ported to KDE Frameworks 5. The problem is that konqueror has been without a full time mantainer for a long time, since the original author shifted his attention to other parts of KDE. Since last year I took up the role of konqueror mantainer, but unfortunately I had much less free time to work on it than I hoped for back then. Currently, the situation regarding konqueror is the following: * the application itself has been ported from KDE4/Qt4 to KDE Frameworks/Qt5 by its original author (David Faure). This means that it compiles and mostly works, including the integration with dolphin. There are several bugs, however. * I've ported most of the plugins (only the user agent plugin doesn't work completely) * I'm in the process of porting the sidebar, which is the last part of konqueror still using KDE 4. However, different from the rest of the application, porting this is going to be a lot of work. This is because the sidebar (or rather, parts of it) still uses a compatibility library used to port applications from KDE 3 to KDE 4. This library, of course, doesn't exist in KDE Frameworks, which means that a large part of the sidebar code needs to be almost rewritten. I've thought of writing an ebuild for the developer version of konqueror, (a one, of course), since the application is currently mostly working (I use it as my primary browser, even if I have to switch to other browser every now and then for sites which won't work with it). The reason I haven't done so is that konqueror requires libkonq and that the 4 and 5 versions of libkonq can't be installed together. Since many applications from kde 4 still use it, this would create an issue. Stefano
