Re: [gentoo-user] Circular dependencies...
Khumba b...@khumba.net [14-01-06 08:32]:
On Mon, 6 Jan 2014 08:18:20 +0100
meino.cra...@gmx.de wrote:
Hi,
I got this:
# emerge --update --newuse --deep --with-bdeps=y @world -vp
These are the packages that would be merged, in order:
Calculating dependencies... done!
[nomerge ] virtual/rubygems-6:ruby20 [1:ruby18, 4:ruby19]
RUBY_TARGETS=(ruby20)
[nomerge ] dev-lang/ruby-2.0.0_p353:2.0 [1.8.7_p374:1.8,
1.9.3_p484:1.9] USE=berkdb doc gdbm ipv6 ncurses rdoc readline ssl -debug
-examples -rubytests -socks5 -tk (-xemacs)
[nomerge ] dev-ruby/json-1.8.0 [1.7.7] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%)
[nomerge ]dev-ruby/rake-0.9.6 [0.9.2.2] USE=doc {-test}
(-bash-completion%) RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby)
(-ree18%)
[nomerge ] dev-ruby/rdoc-4.0.1-r1 [3.12.2] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby)
[ebuild U ] dev-ruby/racc-1.4.9 [1.4.8] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%) 107 kB
[ebuild U ] dev-ruby/json-1.8.0 [1.7.7] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%) 146 kB
[ebuild U ] dev-ruby/rake-0.9.6 [0.9.2.2] USE=doc {-test}
(-bash-completion%) RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby)
(-ree18%) 121 kB
[ebuild U ]dev-ruby/rdoc-4.0.1-r1 [3.12.2] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) 457 kB
Total: 4 packages (4 upgrades), Size of downloads: 830 kB
* Error: circular dependencies:
(dev-ruby/rdoc-4.0.1-r1::gentoo, ebuild scheduled for merge) depends on
(dev-ruby/racc-1.4.9::gentoo, ebuild scheduled for merge) (buildtime)
(dev-ruby/rdoc-4.0.1-r1::gentoo, ebuild scheduled for merge) (buildtime)
It might be possible to break this cycle
by applying the following change:
- dev-ruby/racc-1.4.9 (Change USE: -doc)
I added this to /etc/portage/package.use:
=dev-ruby/racc-1.49 -doc
1.49 != 1.4.9, is that not just a typo in your email?
Cheers,
Khumba
and have set doc in the USE flags in make.conf.
I am still getting the above output...
How can I get out of this cycle out of this cycle out of this cycles
out of... ;)
Best regards,
mcc
Hi Khumba,
oh damn...I didn't saw that! Thank you very much !!!
Now the circle is moved a little further...I got:
[nomerge ] virtual/rubygems-6:ruby20 [1:ruby18, 4:ruby19]
RUBY_TARGETS=(ruby20)
[nomerge ] dev-lang/ruby-2.0.0_p353:2.0 [1.8.7_p374:1.8, 1.9.3_p484:1.9]
USE=berkdb doc gdbm ipv6 ncurses rdoc readline ssl -debug -examples -rubytests
-socks5 -tk (-xemacs)
[nomerge ] dev-ruby/json-1.8.0 [1.7.7] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%)
[nomerge ]dev-ruby/rake-0.9.6 [0.9.2.2] USE=doc {-test}
(-bash-completion%) RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%)
[nomerge ] dev-ruby/rdoc-4.0.1-r1 [3.12.2] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby)
[ebuild U ] dev-ruby/racc-1.4.9 [1.4.8] USE=-doc* {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%) 107 kB
[ebuild U ] dev-ruby/json-1.8.0 [1.7.7] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%) 146 kB
[ebuild U ] dev-ruby/rake-0.9.6 [0.9.2.2] USE=doc {-test}
(-bash-completion%) RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) (-ree18%)
121 kB
[ebuild U ]dev-ruby/rdoc-4.0.1-r1 [3.12.2] USE=doc {-test}
RUBY_TARGETS=ruby18 ruby19 ruby20%* (-jruby) 457 kB
Total: 4 packages (4 upgrades), Size of downloads: 830 kB
* Error: circular dependencies:
(dev-ruby/rake-0.9.6::gentoo, ebuild scheduled for merge) depends on
(dev-ruby/rdoc-4.0.1-r1::gentoo, ebuild scheduled for merge) (buildtime)
(dev-ruby/racc-1.4.9::gentoo, ebuild scheduled for merge) (buildtime)
(dev-ruby/rake-0.9.6::gentoo, ebuild scheduled for merge) (buildtime)
It might be possible to break this cycle
by applying the following change:
- dev-ruby/rake-0.9.6 (Change USE: -doc)
I will now add that to package.use and will see how far it will
take...
Thanks a lot again! :)
Cheers,
mcc
Re: [gentoo-user] Circular dependencies...
On Mon, 6 Jan 2014 08:18:20 +0100, meino.cra...@gmx.de wrote: I added this to /etc/portage/package.use: =dev-ruby/racc-1.49 -doc and have set doc in the USE flags in make.conf. I am still getting the above output... USE=doc is a classic way of getting circular dependencies, usually unnecessarily. You only need USE=doc if you are doing development work with a package, so it should never need to be set globally. The flag does not affect the installation of general man/info pages, except with the odd exception (like ffmpeg). % euses doc doc - Add extra documentation (API, Javadoc, etc). It is recommended to enable per package instead of globally -- Neil Bothwick Diarrhoea is hereditary, it runs in your genes. signature.asc Description: PGP signature
Re: [gentoo-user] Circular dependencies...
Neil Bothwick n...@digimed.co.uk [14-01-06 09:28]: On Mon, 6 Jan 2014 08:18:20 +0100, meino.cra...@gmx.de wrote: I added this to /etc/portage/package.use: =dev-ruby/racc-1.49 -doc and have set doc in the USE flags in make.conf. I am still getting the above output... USE=doc is a classic way of getting circular dependencies, usually unnecessarily. You only need USE=doc if you are doing development work with a package, so it should never need to be set globally. The flag does not affect the installation of general man/info pages, except with the odd exception (like ffmpeg). % euses doc doc - Add extra documentation (API, Javadoc, etc). It is recommended to enable per package instead of globally -- Neil Bothwick Diarrhoea is hereditary, it runs in your genes. Hi Neil, thank you for the inofrmation! :)) I am currently on the way of installing texlive completly and I cannot get enough documentation about it... So I thought, that doc was a good idea...but iw seems not to be... I will switch it off again. On the other side: Not being able to install docs globally seems to be a logical problem (beside being useless...), which needs to be fixed...right? Best regards, mcc
Re: [gentoo-user] Circular dependencies...
On Mon, 6 Jan 2014 09:36:40 +0100, meino.cra...@gmx.de wrote: thank you for the inofrmation! :)) I am currently on the way of installing texlive completly and I cannot get enough documentation about it... So I thought, that doc was a good idea...but iw seems not to be... It may be a good idea, for that package, but not globally. Even so, unless you need API documentation, I doubt you want that flag. I will switch it off again. On the other side: Not being able to install docs globally seems to be a logical problem (beside being useless...), which needs to be fixed...right? The usual solution is to install with -doc first and then enable the flag. This often brings in a long list of dependencies, which is where the problems can arise. I don't have a problems with Gentoo devs not spending time solving a problems that should not occur with sensible use. -- Neil Bothwick On the other hand, you have different fingers. signature.asc Description: PGP signature
Re: [gentoo-user] Confusion about slot conflict
I have just checked the ebuild file of libpng-1.6.8, actually it's is stable for X86. So I am confused why emerge can not upgrade it automatically... 2014/1/5 Alan McKinnon alan.mckin...@gmail.com On 05/01/2014 17:57, 张东亚 wrote: Hi list, When I do a world upgrade, I have encountered the following slot conflict: media-libs/libpng:0 (media-libs/libpng-1.5.17-r1::gentoo, installed) pulled in by media-libs/libpng:0/0= required by (dev-python/wxpython-2.8.12.1-r1::gentoo, installed) =media-libs/libpng-1.4:0/0= required by (net-libs/webkit-gtk-1.8.3-r201::gentoo, installed) media-libs/libpng:0/0= required by (app-text/poppler-0.22.5::gentoo, installed) media-libs/libpng:0/0= required by (media-libs/libwebp-0.3.1::gentoo, installed) media-libs/libpng:0/0= required by (net-print/cups-filters-1.0.36-r1::gentoo, installed) media-libs/libpng:0/0= required by (kde-base/kdelibs-4.11.4::gentoo, installed) (media-libs/libpng-1.6.8::gentoo, ebuild scheduled for merge) pulled in by (no parents that aren't satisfied by other packages in this slot) My question is, seems upgrade to libpng-1.6.8 will solve the conflict, why emerge cannot proceed this automatically? or there are some switches to controller this? Thanks a lot. libpng-1.5.17-r1 is latest stable version libpng-1.6.8 is latest unstable version I suspect you are probably running a stable system with several packages marked unstable and one of those requires libpng-1.6.8 To check, please post what is your ACCEPT_KEYWORDS in make.conf? what is the output of grep -r libpng /etc/portage/ -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Updating $PATH variable permanently for root not working
On 2014-01-05 4:14 PM, Alan McKinnon alan.mckin...@gmail.com wrote: Head spinning yet? Just be done with all that nonsense and do this: Put this line as the only non-comment line in .bash_profile [[ -f ~/.bashrc ]] . ~/.bashrc and put all your shell start-up stuff in .bashrc (moving them out of .bash_profile if necessary This way everything is still unbelievably complex but at least the visible problems mostly just go away Wow... or rather... UGH Anyway, thanks Alan, that did indeed do the trick... :)
Re: [gentoo-user] how to use my SSD the right way ;-)
I consider swapping SSDs ... the Vertex3 from my desktop (replaced by the new Samsung 840 EVO) could replace the even older Intel SSD. I assume that the Vertex3 would be faster than the Intel ... (at least one generation younger). How to migrate? dd isn't recommended, right? rsync the partitions one by one? But one of them is encrypted ... Hints welcome (or even the message that this isn't worth the effort). Stefan
Re: [gentoo-user] how to use my SSD the right way ;-)
On Mon, 06 Jan 2014 16:31:55 +0100, Stefan G. Weichinger wrote: How to migrate? dd isn't recommended, right? rsync the partitions one by one? But one of them is encrypted ... Set up a new encrypted partition and rsync the contents. -- Neil Bothwick Bagpipe for free: Stuff cat under arm. Pull legs, chew tail. signature.asc Description: PGP signature
Re: [gentoo-user] Circular dependencies...
On Mon, 6 Jan 2014 08:57:22 + Neil Bothwick n...@digimed.co.uk wrote: On Mon, 6 Jan 2014 09:36:40 +0100, meino.cra...@gmx.de wrote: thank you for the inofrmation! :)) I am currently on the way of installing texlive completly and I cannot get enough documentation about it... So I thought, that doc was a good idea...but iw seems not to be... It may be a good idea, for that package, but not globally. Even so, unless you need API documentation, I doubt you want that flag. Also it makes the compile time for some packages longer; kdelibs is a case where it takes significantly longer (30+ minutes iirc? it's been a while). Sometimes I think it'd be nice to be able to enable all manuals / user guides but not necessarily API documentation (examples: wireshark, postgresql-server, ghc), but since that's the extent of my examples, and two of them are development-related, I doubt it's worth the effort. - Khumba
Re: [gentoo-user] how to use my SSD the right way ;-)
Am 06.01.2014 16:40, schrieb Neil Bothwick: On Mon, 06 Jan 2014 16:31:55 +0100, Stefan G. Weichinger wrote: How to migrate? dd isn't recommended, right? rsync the partitions one by one? But one of them is encrypted ... Set up a new encrypted partition and rsync the contents. Yep. Thanks ... could have thought of this as well.
Re: [gentoo-user] how to use my SSD the right way ;-)
Am 06.01.2014 17:16, schrieb Stefan G. Weichinger: Am 06.01.2014 16:40, schrieb Neil Bothwick: On Mon, 06 Jan 2014 16:31:55 +0100, Stefan G. Weichinger wrote: How to migrate? dd isn't recommended, right? rsync the partitions one by one? But one of them is encrypted ... Set up a new encrypted partition and rsync the contents. Yep. Thanks ... could have thought of this as well. in terms of speed, it should be worth the effort: http://www.harddrivebenchmark.net/hdd.php?hdd=INTEL+SSDSA2M080G2GC The Vertex3 seems to be way faster.
[gentoo-user] Re: Confusion about slot conflict
On Sun, 5 Jan 2014 15:57:10 + 张东亚 fortitude.zh...@gmail.com wrote: Hi list, When I do a world upgrade, I have encountered the following slot conflict: media-libs/libpng:0 (media-libs/libpng-1.5.17-r1::gentoo, installed) pulled in by media-libs/libpng:0/0= required by (dev-python/wxpython-2.8.12.1-r1::gentoo, installed) =media-libs/libpng-1.4:0/0= required by (net-libs/webkit-gtk-1.8.3-r201::gentoo, installed) media-libs/libpng:0/0= required by (app-text/poppler-0.22.5::gentoo, installed) media-libs/libpng:0/0= required by (media-libs/libwebp-0.3.1::gentoo, installed) media-libs/libpng:0/0= required by (net-print/cups-filters-1.0.36-r1::gentoo, installed) media-libs/libpng:0/0= required by (kde-base/kdelibs-4.11.4::gentoo, installed) (media-libs/libpng-1.6.8::gentoo, ebuild scheduled for merge) pulled in by (no parents that aren't satisfied by other packages in this slot) My question is, seems upgrade to libpng-1.6.8 will solve the conflict, why emerge cannot proceed this automatically? or there are some switches to controller this? Thanks a lot. I had exactly the same weirdness after 1.6.8 was stabilized. I waited a few hours, re-synced the tree, and then updated without any warnings about blocks; the old libpng was unmerged and 1.6.8 was merged automagically as one would expect. I don't know what caused the problem or what fixed it.
[gentoo-user] GSOC discussion
Google, Summer of Code, is a very wonderful idea for the Open Source Community: http://www.google-melange.com/gsoc/homepage/google/gsoc2014 No doubt many are scheming as to what would be good project ideas. Many really good ideas come from the rank and file of users. (Gmane.org is back online). Those (gifted? fledgling_gifted) often code good ideas directly. I think now is the time to have a robust (community) discussion on what the users of this list think would be good ideas for GSOC projects, which are Gentoo centric. My experiences with ALL current search engines, is there is quite a lot of low_quality results; which may be a direct result of the masses of those folks trying to learn/use unix/linux/gentoo by poorly formed google (et. al.) searches. Those low quality efforts statistically stack up against those few robustly accurate searches (from folks like the users of this list) so as to contaiminate search engines by the sheer number of poorly formed search engine efforts. I.E. it's not the fault of the search engine(?). So, Naturally, here is my idea. Collectively, if you look at this list archive and other such gentoo-centric resources, there is a wealth of good information that traverses our paths, and is then subsequently burried into the annals of lifeless data, only occasionally pinged by exasperating google searches, or gleaned from the multitude of records we each privately maintain. So, some sort of tool/app/parser that can glean information from the various gentoo-centric resources and store the data for later searching would be useful. Then a variety of experimental front ends could be tested (or developed anew) to query that data with relevant searches and very targeted searches. By limiting the focus to all things Gentoo, we could serve ourselves up crudely basic, but highly valuable information specific to the Gentoo-centric paramenters that one uses for this in-house database. This effort would mostly priortize the data returned in searches by some sort of *quality metric*. For example it could be simply that the most experienced dev would rate a *100 (times 100) multiplier on the quality of their searches. Recognized power users might rate a *25 rating. Ordinary folks a *1 rating on quality. This system would not even have to be open to the rank and file internet, except as a source of data returned to search engines like google, for example. Other distros could eventually do the same thing, and then Google could probe these databases, gaining specific user-community preferences of excellent quality (accuracy) search-data from folks steeped in all things Gentoo. Basically, a schemea is developed that allows the strongest of the Gentoo community to radically improve the quality of gentoo-specific searches, with very little efforts. For the devs and such, it could merely be a front-end to their own google searches. Folks would not have to use this front-end, if they do not want to. So, what is your idea for GSOC_2014? http://wiki.gentoo.org/index.php?title=Google_Summer_of_Code/2013/Ideas
[gentoo-user] dovecot + inotify max_user instances
Hello,
Last year Timo (the dovecot author) remoted into our server and migrated
us from courier-imap to dovecot. One of the things we ran into
afterwards was this warning in the logs:
Warning: Inotify instance limit for user 200 (UID vmail) exceeded, disabling.
Increase /proc/sys/fs/inotify/max_user_instances
He solved it by doing:
echo 1024 /proc/sys/fs/inotify/max_user_instances
Then he said he was going to solve this permanently by changing the init
script...
Well, my shiny new server encountered the same issue once it went under
load this morning, so I had to dig out this solution - but obviously
this will come back again the next time I reboot, so I want to find a
permanent solution.
I guess I could manually modify my init script, but I'm clueless as to
where or how.
If someone can help me figure out how to do this, I'll go open a bug to
get this added to the gentoo init script permanently.
The current (gentoo) init script is:
#!/sbin/runscript
# Copyright 1999-2012 Gentoo Foundation
# Distributed under the terms of the GNU General Public License, v2 or
later
# $Header:
/var/cvsroot/gentoo-x86/net-mail/dovecot/files/dovecot.init-r4,v 1.1
2012/12/01 10:40:49 eras Exp $
extra_started_commands=reload
depend() {
need localmount
before postfix
after bootmisc ldap mysql ntp-client ntpd postgresql saslauthd
slapd
use logger net
}
checkconfig() {
DOVECOT_INSTANCE=${SVCNAME##*.}
if [ -n ${DOVECOT_INSTANCE} -a ${SVCNAME} != dovecot ]; then
DOVECOT_CONF=/etc/dovecot/dovecot.${DOVECOT_INSTANCE}.conf
else
DOVECOT_CONF=/etc/dovecot/dovecot.conf
fi
if [ ! -e ${DOVECOT_CONF} ]; then
eerror You will need an ${DOVECOT_CONF} first
return 1
fi
if [ -x /usr/sbin/dovecot ]; then
DOVECOT_BASEDIR=$(/usr/sbin/dovecot -c ${DOVECOT_CONF}
-a | grep '^base_dir = ' | sed 's/^base_dir = //')
else
eerror dovecot not executable
return 1
fi
DOVECOT_BASEDIR=${DOVECOT_BASEDIR:-/var/run/dovecot}
DOVECOT_PIDFILE=${DOVECOT_BASEDIR}/master.pid
}
start() {
checkconfig || return 1
ebegin Starting ${SVCNAME}
start-stop-daemon --start --exec /usr/sbin/dovecot \
--pidfile ${DOVECOT_PIDFILE} -- -c ${DOVECOT_CONF}
eend $?
}
stop() {
checkconfig || return 1
ebegin Stopping ${SVCNAME}
start-stop-daemon --stop --exec /usr/sbin/dovecot \
--pidfile ${DOVECOT_PIDFILE}
eend $?
}
reload() {
checkconfig || return 1
ebegin Reloading ${SVCNAME} configs and restarting auth/login
processes
start-stop-daemon --signal HUP --exec /usr/sbin/dovecot \
--pidfile ${DOVECOT_PIDFILE}
eend $?
}
Thx for any help...
Charles
Re: [gentoo-user] dovecot + inotify max_user instances
On 06/01/2014 20:50, Tanstaafl wrote: Hello, Last year Timo (the dovecot author) remoted into our server and migrated us from courier-imap to dovecot. One of the things we ran into afterwards was this warning in the logs: Warning: Inotify instance limit for user 200 (UID vmail) exceeded, disabling. Increase /proc/sys/fs/inotify/max_user_instances He solved it by doing: echo 1024 /proc/sys/fs/inotify/max_user_instances Then he said he was going to solve this permanently by changing the init script... Well, my shiny new server encountered the same issue once it went under load this morning, so I had to dig out this solution - but obviously this will come back again the next time I reboot, so I want to find a permanent solution. I guess I could manually modify my init script, but I'm clueless as to where or how. If someone can help me figure out how to do this, I'll go open a bug to get this added to the gentoo init script permanently. [snip] There's a more generic solution that doesn't involve hacking the init script. Not all dovecot users will appreciate such a change be put on their systems and that kind of config is best left up to the individual sysadmin to determine what suits them best (no single value satisfies everyone). Instead, add that echo statement to a file in /etc/local.d/ and add the local service to your default runlevel. It will tweak the max_user_instances knob for you when it runs, same with all such local knob fiddling you do to optimize performance. I think overall this is better than hacking around in dovecot's start-up script (a rather unsuitable place for it imho) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Re: Confusion about slot conflict
On 06/01/2014 19:58, »Q« wrote: On Sun, 5 Jan 2014 15:57:10 + 张东亚 fortitude.zh...@gmail.com wrote: Hi list, When I do a world upgrade, I have encountered the following slot conflict: media-libs/libpng:0 (media-libs/libpng-1.5.17-r1::gentoo, installed) pulled in by media-libs/libpng:0/0= required by (dev-python/wxpython-2.8.12.1-r1::gentoo, installed) =media-libs/libpng-1.4:0/0= required by (net-libs/webkit-gtk-1.8.3-r201::gentoo, installed) media-libs/libpng:0/0= required by (app-text/poppler-0.22.5::gentoo, installed) media-libs/libpng:0/0= required by (media-libs/libwebp-0.3.1::gentoo, installed) media-libs/libpng:0/0= required by (net-print/cups-filters-1.0.36-r1::gentoo, installed) media-libs/libpng:0/0= required by (kde-base/kdelibs-4.11.4::gentoo, installed) (media-libs/libpng-1.6.8::gentoo, ebuild scheduled for merge) pulled in by (no parents that aren't satisfied by other packages in this slot) My question is, seems upgrade to libpng-1.6.8 will solve the conflict, why emerge cannot proceed this automatically? or there are some switches to controller this? Thanks a lot. I had exactly the same weirdness after 1.6.8 was stabilized. I waited a few hours, re-synced the tree, and then updated without any warnings about blocks; the old libpng was unmerged and 1.6.8 was merged automagically as one would expect. I don't know what caused the problem or what fixed it. Ah, that explains it. I last synced 3 days ago Obviously, updating libpng in the tree was a two stage process, with a longer than expected delay between them. Or maybe a bug the dev didn't anticipate showed up right away and needed fixing. You and fortitude both sync'ed in this middle period and got hit by the bug. It happens a lot actually :-) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] dovecot + inotify max_user instances
Alan McKinnon alan.mckin...@gmail.com wrote: On 06/01/2014 20:50, Tanstaafl wrote: Hello, Last year Timo (the dovecot author) remoted into our server and migrated us from courier-imap to dovecot. One of the things we ran into afterwards was this warning in the logs: Warning: Inotify instance limit for user 200 (UID vmail) exceeded, disabling. Increase /proc/sys/fs/inotify/max_user_instances He solved it by doing: echo 1024 /proc/sys/fs/inotify/max_user_instances Then he said he was going to solve this permanently by changing the init script... Well, my shiny new server encountered the same issue once it went under load this morning, so I had to dig out this solution - but obviously this will come back again the next time I reboot, so I want to find a permanent solution. I guess I could manually modify my init script, but I'm clueless as to where or how. If someone can help me figure out how to do this, I'll go open a bug to get this added to the gentoo init script permanently. [snip] There's a more generic solution that doesn't involve hacking the init script. Not all dovecot users will appreciate such a change be put on their systems and that kind of config is best left up to the individual sysadmin to determine what suits them best (no single value satisfies everyone). Instead, add that echo statement to a file in /etc/local.d/ and add the local service to your default runlevel. It will tweak the max_user_instances knob for you when it runs, same with all such local knob fiddling you do to optimize performance. I think overall this is better than hacking around in dovecot's start-up script (a rather unsuitable place for it imho) I put such a thing in /etc/sysctl.conf like this -- I don't have dovecot, but I needed it for crashplan fs.inotify.max_user_watches=100 or whatever value suits. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] Circular dependencies...
On Mon, 6 Jan 2014 08:03:42 -0800, Khumba wrote: Sometimes I think it'd be nice to be able to enable all manuals / user guides but not necessarily API documentation That's how USE=-doc is supposed to work. Man pages etc, should be installed by default unless FEATURES=nodoc noman noinfo is set. -- Neil Bothwick Unsolicited advice is the junk mail of life signature.asc Description: PGP signature
Re: [gentoo-user] dovecot + inotify max_user instances
On 06/01/2014 21:53, cov...@ccs.covici.com wrote: Instead, add that echo statement to a file in /etc/local.d/ and add the local service to your default runlevel. It will tweak the max_user_instances knob for you when it runs, same with all such local knob fiddling you do to optimize performance. I think overall this is better than hacking around in dovecot's start-up script (a rather unsuitable place for it imho) I put such a thing in /etc/sysctl.conf like this -- I don't have dovecot, but I needed it for crashplan fs.inotify.max_user_watches=100 or whatever value suits. /me bangs head Yes of course, that is a better solution for things with a sysctl. I look at the echo statement and it writes to something in /proc My brain said it writes to something in /sys Now, where do I leave those memory-booster pills for old people again? -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Updating $PATH variable permanently for root not working
On 06/01/2014 00:05, Michael Orlitzky wrote:
On 01/05/2014 04:14 PM, Alan McKinnon wrote:
This way everything is still unbelievably complex but at least the
visible problems mostly just go away
There is an apparently empty directory, /etc/skel, that upon closer
inspection contains some nice default bash junk:
$ ls -a /etc/skel/
total 32K
drwxr-xr-x 3 root root 4.0K 2013-06-06 10:53 .
drwxr-xr-x 113 root root 12K 2014-01-05 01:24 ..
-rw-r--r-- 1 root root 127 2013-06-06 10:53 .bash_logout
-rw-r--r-- 1 root root 193 2013-06-06 10:53 .bash_profile
-rw-r--r-- 1 root root 551 2013-06-06 10:53 .bashrc
drwx-- 2 root root 4.0K 2007-11-23 14:25 .ssh
The 'useradd' program {might,should} install these for you; if not it
can be coaxed into it with the --skel flag.
The .bash_profile in there does what Alan suggests.
Ah, but there's a snag with the root account which is what Charles is using
useradd root is never run on a Unix boot and /etc/skel is only copied
over by using useradd. The root account is created at install time
unpacking the stage 3 IIRC so the profiles it gets are whatever is in
the tarball.
To get /etc/skel for root, one has to copy the file.
I keep forgetting to do this myself on installs, this the main reason
why I know more about bash startup than I should :-)
--
Alan McKinnon
alan.mckin...@gmail.com
Re: [gentoo-user] Updating $PATH variable permanently for root not working
On 06/01/2014 14:28, Tanstaafl wrote: On 2014-01-05 4:14 PM, Alan McKinnon alan.mckin...@gmail.com wrote: Head spinning yet? Just be done with all that nonsense and do this: Put this line as the only non-comment line in .bash_profile [[ -f ~/.bashrc ]] . ~/.bashrc and put all your shell start-up stuff in .bashrc (moving them out of .bash_profile if necessary This way everything is still unbelievably complex but at least the visible problems mostly just go away Wow... or rather... UGH Anyway, thanks Alan, that did indeed do the trick... :) You're welcome. I figured you deserved to have the whole ugly truth for your reading pleasure, with the option of a tl;dr of course ;-) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] dovecot + inotify max_user instances
On 2014-01-06 2:32 PM, Alan McKinnon alan.mckin...@gmail.com wrote: I think overall this is better than hacking around in dovecot's start-up script (a rather unsuitable place for it imho Actually, a follow-up from Timo said he was planning on making this a configuration setting (default blank), and the init script would then just use this setting (if set). Which probably makes the most sense...
Re: [gentoo-user] Do I really need a sshd?
On Sun, 5 Jan 2014 10:16:20 + Neil Bothwick n...@digimed.co.uk wrote: On Sun, 5 Jan 2014 11:36:20 +0200, Gevisz wrote: I was afraid to run etc-update as man says it will replace everything automatically. However, I run dispatch-conf and it does not see any problems at /etc/ssh, which have only the following three files: moduli, ssh_config, sshd_config (though I have added /etc/ssh to CONFIG_PROTECT_MASK). Why did you do that? By masking out config file protection for /etc/ssh there will never be anything to be managed by etc-update as you have told portage to replace those files blindly and without asking. From man dispatch-conf: CONFIG_PROTECT_MASK is a make.conf setting, read that man page. It means your config files are overwritten at install time, way be for you run dispatch-conf or one of its friends. The man page for make.conf actually refers to the CONFIGURATION FILES section of emerge man page. There, indeed, everything explained very clearly and in details. However, it does not work so for my system. Thus, I have the following settings $ echo $CONFIG_PROTECT /usr/share/gnupg/qualified.txt $ echo $CONFIG_PROTECT_MASK /etc/gentoo-release /etc/sandbox.d /etc/fonts/fonts.conf /etc/terminfo /etc/dconf /etc/ca-certificates.conf /etc/texmf/web2c /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/revdep-rebuild but, nevertheless, I see the usual protected behavior for all config files in my /etc directory and its subdirectories, that is, instead of overriding them while running emerge world, the system creates the corresponding ._cfg_* files. And this, in my view, contradicts with the following statement from emerge man pages: Protected directories are set using the CONFIG_PROTECT variable, as in my case /etc have not been set in my CONFIG_PROTECT variable. dispatch-conf will check all directories in the CONFIG_PROTECT variable. All config files found in CONFIG_PROTECT_MASK will automatically be updated for you by dispatch-conf. 3) I will continue to do this job manually with gvimdiff as I have found it much more convenient than dispatch-conf (gvimdiff shows the differences a way much better). I prefer conf-update but most of these tools allow you to specify your own diff program if you don't like the default. I use colordiff with conf-update. Ok, thank you. I will try some of these tools together with gvimdiff when I will have to update my config files in the future.
Re: [gentoo-user] Do I really need a sshd?
On Sun, 05 Jan 2014 14:51:21 +0100 Volker Armin Hemmann volkerar...@googlemail.com wrote: Am 05.01.2014 11:04, schrieb Peter Humphrey: On Sunday 05 Jan 2014 11:36:20 Gevisz wrote: From man dispatch-conf: dispatch-conf will check all directories in the CONFIG_PROTECT variable. All config files found in CONFIG_PROTECT_MASK will automatically be updated for you by dispatch-conf. Have you tried another updater of config files? I still use the basic etc- update, but there are also app-portage/cfg-update and app-portage/conf-update. You might feel happier with one of those three. I am using cfg-update for years without problems whatsoever. Thank you, I will try it once more.
Re: [gentoo-user] Do I really need a sshd?
On Sun, 05 Jan 2014 10:04:54 + Peter Humphrey pe...@prh.myzen.co.uk wrote: On Sunday 05 Jan 2014 11:36:20 Gevisz wrote: From man dispatch-conf: dispatch-conf will check all directories in the CONFIG_PROTECT variable. All config files found in CONFIG_PROTECT_MASK will automatically be updated for you by dispatch-conf. Have you tried another updater of config files? I still use the basic etc- update, but there are also app-portage/cfg-update and app-portage/conf-update. You might feel happier with one of those three. Thank you, I will try them all when I will have to update my config files.
Re: [gentoo-user] dovecot + inotify max_user instances
On 06/01/2014 23:04, Tanstaafl wrote: On 2014-01-06 2:32 PM, Alan McKinnon alan.mckin...@gmail.com wrote: I think overall this is better than hacking around in dovecot's start-up script (a rather unsuitable place for it imho Actually, a follow-up from Timo said he was planning on making this a configuration setting (default blank), and the init script would then just use this setting (if set). Which probably makes the most sense... Personally, I'd put it in sysctls per covics post. Can you imagine three years hence trying to find why max_user_instances is set non-default and *nothing* in the usual places gives a clue why? When you eventually resort to grep -r max_user_instances /etc and realise that a kernel setting is being changed in your imap server config, you will curse and swear to the heavens (been there, done that) Of course, it's up to you. I just like to make config settings in places that are the most obvious to everyone who is not me -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Do I really need a sshd?
On Mon, 6 Jan 2014 23:50:02 +0200, Gevisz wrote: On Sun, 5 Jan 2014 10:16:20 + Neil Bothwick n...@digimed.co.uk wrote: CONFIG_PROTECT_MASK is a make.conf setting, read that man page. It means your config files are overwritten at install time, way be for you run dispatch-conf or one of its friends. The man page for make.conf actually refers to the CONFIGURATION FILES section of emerge man page. There, indeed, everything explained very clearly and in details. However, it does not work so for my system. Thus, I have the following settings $ echo $CONFIG_PROTECT /usr/share/gnupg/qualified.txt CONFIG_PROTECT is not a global environment variable. Also, variables like this are not only set in make.conf but in profiles and even on the command line. Read man portage for more detail. The correct way to see what settings are in place is to run emerge --info emerge --info | grep CONFIG_PROTECT And this, in my view, contradicts with the following statement from emerge man pages: Protected directories are set using the CONFIG_PROTECT variable, as in my case /etc have not been set in my CONFIG_PROTECT variable. See above. /etc is protected in the base profile, inherited by all others. -- Neil Bothwick Logic will get you from A to B. Imagination will take you everywhere. (Albert Einstein) signature.asc Description: PGP signature
Re: [gentoo-user] dovecot + inotify max_user instances
On 07/01/14 02:50, Tanstaafl wrote: Hello, Last year Timo (the dovecot author) remoted into our server and migrated us from courier-imap to dovecot. One of the things we ran into afterwards was this warning in the logs: Warning: Inotify instance limit for user 200 (UID vmail) exceeded, disabling. Increase /proc/sys/fs/inotify/max_user_instances He solved it by doing: echo 1024 /proc/sys/fs/inotify/max_user_instances Then he said he was going to solve this permanently by changing the init script... Why initscript? - as a user you should put this in a /etc/local.d script. Leave the system files as they are. BillK
