[gentoo-user] Re: Cleaning out my world file
On Friday 15 August 2008, Albert Hopkins wrote: On Fri, 2008-08-15 at 08:38 -0500, Dale wrote: Alan McKinnon wrote: On Friday 15 August 2008 14:36:58 Dale wrote: Somewhat still on the same subject since I am still cleaning. Anyway to clean out unneeded files in /etc? I'm thinking about files that may be there but the programs are no longer installed. I read the man page for dep but didn't see anything. Dang thing does a lot tho. You could use the very long way round, something based on this: find /etc/ -type f -exec equery belongs {} \; then leave it alone for an hour or three H, I had to stop that after a few minutes. It sort of took away from my folding. Pushed my CPU to about 80% or so. There has to be a tool for this too. Gentoo has about everything else. I do a similar thing every month as a cron job. It' runs at night so I just get an email the next day. -- #!/bin/bash # Print out orphan files in specified directories find /etc -xdev -type f -print|xargs qfile -o find /usr -xdev \( -path /usr/src -prune \) -o -type f -not -name '*.pyc' \ -not -name '*.pyo' -not -name .keep -print | \ xargs qfile -o find /lib -xdev \( -path /lib/modules -prune \) -o -type f |xargs qfile -o It doesn't handle filenames with spaces, I tried with find /etc -xdev -type f -exec qfile -o {} \; that works with spaces, but has awful performances, so that's completely unusable on /usr. BTW I also tried to remove false positive hiding /usr/portage and the overlays folder . Does anyone know how to deal with spaces (to avoid false positives) and keep reasonable running times? TIA Francesco -- Linux Version 2.6.26-gentoo, Compiled #2 PREEMPT Sat Aug 9 20:21:11 CEST 2008 One 1GHz AMD Athlon 64 Processor, 2GB RAM, 2004.04 Bogomips Total aemaeth
[gentoo-user] e2fsprogs and blocking
Hi! After last syncing 'emerge -pvDuN world' ends with messages shown below. How to resolve this conflict *safely*? I don't want to experiment with fs-related packages :-) Andrew = [ebuild N] sys-libs/e2fsprogs-libs-1.41.0 USE=nls 476 kB [ebuild U ] sys-fs/e2fsprogs-1.41.0 [1.40.11] USE=nls (-static%) 4,161 kB [blocks B ] sys-fs/e2fsprogs-1.41 (is blocking sys-libs/e2fsprogs-libs-1.41.0) [blocks B ] sys-libs/ss (is blocking sys-libs/e2fsprogs-libs-1.41.0) [blocks B ] sys-libs/com_err (is blocking sys-libs/e2fsprogs-libs-1.41.0)
Re: [gentoo-user] Cleaning out my world file
On Sat, 16 Aug 2008 23:55:42 -0500, Dale wrote: Even if this script said something belonged to nothing installed, I would still check to make sure it was safe to delete. That is what the script does, qfile -o file only outputs anything if file does not belong to a package. -- Neil Bothwick Q. Why did the koala fall out of the tree? A. It was dead. signature.asc Description: PGP signature
Re: [gentoo-user] df and du difference
On Sat, 16 Aug 2008 10:20:49 +0430, [EMAIL PROTECTED] wrote: the difference between du and df is about 640 - 188 = 452 MB. and df is showing that my root is full 2.4 times more than du. There is another possibility that no one seems t have mentioned. You may have files hidden inside a mount point that du is unable to see. To test mkdir /mnt/tmp mount --bind / /mnt/tmp then run your du commands over /mnt/tmp. -- Neil Bothwick There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies. The first method is far more difficult -C.A.R. Hoare Neil Bothwick signature.asc Description: PGP signature
Re: [gentoo-user] e2fsprogs and blocking
Am Sonntag, 17. August 2008 09:52:57 schrieb Andrew Gaydenko: After last syncing 'emerge -pvDuN world' ends with messages shown below. How to resolve this conflict *safely*? I don't want to experiment with fs-related packages :-) What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs You might want to run a revdep-rebuild after emerging the new e2fsprogs, although the only thing that happened with libraries AFAICT is that sys-libs/ss and sys-libs/com_err now come combined in the sys-libs/e2fsprogs-libs package without any versioning change, and so revdep-rebuild should find nothing. HTH! --- Heiko.
Re: [gentoo-user] Cleaning out my world file
Neil Bothwick wrote: On Sat, 16 Aug 2008 23:55:42 -0500, Dale wrote: Even if this script said something belonged to nothing installed, I would still check to make sure it was safe to delete. That is what the script does, qfile -o file only outputs anything if file does not belong to a package. Hi Neil, WOW! That would be one HUGE list. It's something like 4,000 of them. O_O Might be quicker to reinstall. LOL This install is about 5 years old now. One reason I wanted to do a little, key word little, cleaning. Dale :-) :-)
Re: [gentoo-user] e2fsprogs and blocking
=== On Sunday 17 August 2008, Heiko Wundram wrote: === Am Sonntag, 17. August 2008 09:52:57 schrieb Andrew Gaydenko: After last syncing 'emerge -pvDuN world' ends with messages shown below. How to resolve this conflict *safely*? I don't want to experiment with fs-related packages :-) What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs Heiko, Why 'oneshot' option is used? Andrew You might want to run a revdep-rebuild after emerging the new e2fsprogs, although the only thing that happened with libraries AFAICT is that sys-libs/ss and sys-libs/com_err now come combined in the sys-libs/e2fsprogs-libs package without any versioning change, and so revdep-rebuild should find nothing. HTH! --- Heiko.
Re: [gentoo-user] e2fsprogs and blocking
Am Sonntag, 17. August 2008 10:08:46 schrieb Andrew Gaydenko: === On Sunday 17 August 2008, Heiko Wundram wrote: === Am Sonntag, 17. August 2008 09:52:57 schrieb Andrew Gaydenko: After last syncing 'emerge -pvDuN world' ends with messages shown below. How to resolve this conflict *safely*? I don't want to experiment with fs-related packages :-) What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs Why 'oneshot' option is used? So that you don't record sys-fs/e2fsprogs in the world set. It's pulled in as a dependency from the system set anyway (that's why the --unmerge will tell you that you're unmerging a system package, and that you'll potentially break your system), and I personally don't like unnecessary packages cluttering up my world set, so I always use --oneshot in cases like this. But, if you leave it out, it does no harm, either. --- Heiko.
Re: [gentoo-user] e2fsprogs and blocking
Heiko Wundram [EMAIL PROTECTED] writes: What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs Be *very* careful about doing that. wget (amongst other packages) uses libcom_err, so you will not be able to fetch any packages (including e2fsprogs-lib) if you follow the above instructions. See bug https://bugs.gentoo.org/show_bug.cgi?id=234907 for details and for a safer workaround.
Re: [gentoo-user] e2fsprogs and blocking
=== On Sunday 17 August 2008, Heiko Wundram wrote: === ... Why 'oneshot' option is used? So that you don't record sys-fs/e2fsprogs in the world set. It's pulled in as a dependency from the system set anyway (that's why the --unmerge will tell you that you're unmerging a system package, and that you'll potentially break your system), and I personally don't like unnecessary packages cluttering up my world set, so I always use --oneshot in cases like this. But, if you leave it out, it does no harm, either. --- Heiko. Done. Thanks! Andrew
Re: [gentoo-user] e2fsprogs and blocking
=== On Sunday 17 August 2008, Graham Murray wrote: === Heiko Wundram [EMAIL PROTECTED] writes: What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs Be *very* careful about doing that. wget (amongst other packages) uses libcom_err, so you will not be able to fetch any packages (including e2fsprogs-lib) if you follow the above instructions. See bug https://bugs.gentoo.org/show_bug.cgi?id=234907 for details and for a safer workaround. I see, thanks. Fortunately I always fetch all needed packages *before* any updating/installing. Andrew
Re: [gentoo-user] mov to dvd
On Saturday 16 August 2008, Neil Bothwick wrote: On Sat, 16 Aug 2008 10:38:31 +0100, Mick wrote: Is there a way to convert a .mov file to be playable by a dvd player? media-video/tovid Thanks Neil! :) -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Duplicate ca certs
Hi All, I am getting mixed up with update-ca-certificates. It reports that I have duplicates: = # update-ca-certificates Updating certificates in /etc/ssl/certsWARNING: SPI_CA_2006-cacert.pem does not contain a certificate or CRL: skipping WARNING: Verisign_Class_1_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: cacert.org.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate QuoVadis_Root_CA.pem WARNING: Skipping duplicate certificate Verisign_RSA_Secure_Server_CA.pem WARNING: Skipping duplicate certificate America_Online_Root_Certification_Authority_1.pem WARNING: Skipping duplicate certificate America_Online_Root_Certification_Authority_2.pem WARNING: Verisign_Class_3_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate thawteCb.pem WARNING: Skipping duplicate certificate Wells_Fargo_Root_CA.pem WARNING: Skipping duplicate certificate thawteCp.pem WARNING: Skipping duplicate certificate vsign3.pem WARNING: spi-ca.pem does not contain a certificate or CRL: skipping WARNING: Verisign_Secure_Server_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate aoltw1.pem WARNING: Skipping duplicate certificate aoltw2.pem WARNING: Verisign_Class_2_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping done. Running hooks in /etc/ca-certificates/update.ddone. = However, when I check for e.g. vsign3.pem I see this: # ls -la /etc/ssl/certs/vsign* -rw-r--r-- 1 root root 984 Jun 1 09:43 /etc/ssl/certs/vsign1.pem -rw-r--r-- 1 root root 989 Dec 4 2005 /etc/ssl/certs/vsign2.pem -rw-r--r-- 1 root root 984 Jun 1 09:43 /etc/ssl/certs/vsign3.pem -rw-r--r-- 1 root root 976 Jun 1 09:43 /etc/ssl/certs/vsignss.pem -rw-r--r-- 1 root root 1084 Dec 4 2005 /etc/ssl/certs/vsigntca.pem Also, what should I do with those that report does not contain a certificate or CRL: skipping? -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Cleaning out my world file
On Sunday 17 August 2008 06:55:42 Dale wrote: Even if this script said something belonged to nothing installed, I would still check to make sure it was safe to delete. I also keep backups of not only my whole system but also a separate backup of /etc, just in case I edit something badly. I feel that searching for orphaned directories is a better idea and gives a better success. The few orphaned files sitting in /etc/ and unused don't take up that much space. But /usr/share/doc/something is a bit of a warning flag. Find those dirs, judicious use of locate, and you get a much better hit rate -- alan dot mckinnon at gmail dot com
Re: [gentoo-user] e2fsprogs and blocking
Am Sonntag, 17. August 2008 10:40:38 schrieb Graham Murray: Heiko Wundram [EMAIL PROTECTED] writes: What worked fine for me: emerge --unmerge sys-libs/ss sys-libs/com_err sys-fs/e2fsprogs emerge --oneshot -v sys-fs/e2fsprogs Be *very* careful about doing that. wget (amongst other packages) uses libcom_err, so you will not be able to fetch any packages (including e2fsprogs-lib) if you follow the above instructions. See bug https://bugs.gentoo.org/show_bug.cgi?id=234907 for details and for a safer workaround. I cannot reproduce this here: [EMAIL PROTECTED] ~ % ldd /usr/bin/wget linux-gate.so.1 = (0xb8088000) libssl.so.0.9.8 = /usr/lib/libssl.so.0.9.8 (0xb8025000) libcrypto.so.0.9.8 = /usr/lib/libcrypto.so.0.9.8 (0xb7ee6000) libdl.so.2 = /lib/libdl.so.2 (0xb7ee2000) librt.so.1 = /lib/librt.so.1 (0xb7ed9000) libc.so.6 = /lib/libc.so.6 (0xb7d99000) /lib/ld-linux.so.2 (0xb8089000) libpthread.so.0 = /lib/libpthread.so.0 (0xb7d81000) [EMAIL PROTECTED] ~ % None of the specified libraries indirectly references libcom_err, either, and running wget with libcom_err temporarily not present also works, and I did not download the packages before using the commands I noted. Possibly, there's a difference between the x86 wget and the ~x86 wget (I run the latter). --- Heiko.
Re: [gentoo-user] e2fsprogs and blocking
Heiko Wundram [EMAIL PROTECTED] writes: I cannot reproduce this here: [EMAIL PROTECTED] ~ % ldd /usr/bin/wget linux-gate.so.1 = (0xb8088000) libssl.so.0.9.8 = /usr/lib/libssl.so.0.9.8 (0xb8025000) libcrypto.so.0.9.8 = /usr/lib/libcrypto.so.0.9.8 (0xb7ee6000) libdl.so.2 = /lib/libdl.so.2 (0xb7ee2000) librt.so.1 = /lib/librt.so.1 (0xb7ed9000) libc.so.6 = /lib/libc.so.6 (0xb7d99000) /lib/ld-linux.so.2 (0xb8089000) libpthread.so.0 = /lib/libpthread.so.0 (0xb7d81000) [EMAIL PROTECTED] ~ % None of the specified libraries indirectly references libcom_err, either, and running wget with libcom_err temporarily not present also works, and I did not download the packages before using the commands I noted. Possibly, there's a difference between the x86 wget and the ~x86 wget (I run the latter). The difference is the kerberos USE flag which is set in /usr/portage/profiles/targets/desktop/make.defaults
Re: [gentoo-user] Duplicate ca certs
On Sun, 2008-08-17 at 10:39 +0100, Mick wrote: Hi All, I am getting mixed up with update-ca-certificates. It reports that I have duplicates: = # update-ca-certificates Updating certificates in /etc/ssl/certsWARNING: SPI_CA_2006-cacert.pem does not contain a certificate or CRL: skipping WARNING: Verisign_Class_1_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: cacert.org.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate QuoVadis_Root_CA.pem WARNING: Skipping duplicate certificate Verisign_RSA_Secure_Server_CA.pem WARNING: Skipping duplicate certificate America_Online_Root_Certification_Authority_1.pem WARNING: Skipping duplicate certificate America_Online_Root_Certification_Authority_2.pem WARNING: Verisign_Class_3_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate thawteCb.pem WARNING: Skipping duplicate certificate Wells_Fargo_Root_CA.pem WARNING: Skipping duplicate certificate thawteCp.pem WARNING: Skipping duplicate certificate vsign3.pem WARNING: spi-ca.pem does not contain a certificate or CRL: skipping WARNING: Verisign_Secure_Server_OCSP_Responder.pem does not contain a certificate or CRL: skipping WARNING: Skipping duplicate certificate aoltw1.pem WARNING: Skipping duplicate certificate aoltw2.pem WARNING: Verisign_Class_2_Public_Primary_OCSP_Responder.pem does not contain a certificate or CRL: skipping done. Running hooks in /etc/ca-certificates/update.ddone. = However, when I check for e.g. vsign3.pem I see this: # ls -la /etc/ssl/certs/vsign* -rw-r--r-- 1 root root 984 Jun 1 09:43 /etc/ssl/certs/vsign1.pem -rw-r--r-- 1 root root 989 Dec 4 2005 /etc/ssl/certs/vsign2.pem -rw-r--r-- 1 root root 984 Jun 1 09:43 /etc/ssl/certs/vsign3.pem -rw-r--r-- 1 root root 976 Jun 1 09:43 /etc/ssl/certs/vsignss.pem -rw-r--r-- 1 root root 1084 Dec 4 2005 /etc/ssl/certs/vsigntca.pem Also, what should I do with those that report does not contain a certificate or CRL: skipping? When you updated the ca-certificates, you should have gotten a postinst message about broken symlinks that you need to remove.
Re: [gentoo-user] e2fsprogs and blocking
-BEGIN PGP SIGNED MESSAGE- Hash: SHA512 Andrew Gaydenko wrote: Hi! After last syncing 'emerge -pvDuN world' ends with messages shown below. How to resolve this conflict *safely*? I don't want to experiment with fs-related packages :-) Andrew = [ebuild N] sys-libs/e2fsprogs-libs-1.41.0 USE=nls 476 kB [ebuild U ] sys-fs/e2fsprogs-1.41.0 [1.40.11] USE=nls (-static%) 4,161 kB [blocks B ] sys-fs/e2fsprogs-1.41 (is blocking sys-libs/e2fsprogs-libs-1.41.0) [blocks B ] sys-libs/ss (is blocking sys-libs/e2fsprogs-libs-1.41.0) [blocks B ] sys-libs/com_err (is blocking sys-libs/e2fsprogs-libs-1.41.0) The best way to resolve this, I've found is to put the following into your /etc/portage/package.mask file. =sys-libs/e2fsprogs-libs-1.41.0 =sys-fs/e2fsprogs-1.41.0 I masked them because some things depend on com_err, and that is part of e2fsprogs-libs in the 1.41.0 version. Until the upstreams of these packages update their packages, you can't use the new version, unless you're willing to unmerge all of the affected packages. Regards, Chris -BEGIN PGP SIGNATURE- iQIcBAEBCgAGBQJIqBSGAAoJEIAhA8M9p9DAzpkP/jK/Fb8oOpHuT4lb2VNzOlEd fy/vFlqdJ+uMvpD2YtQCrwkBbzIEnyq4oqlYjixH2DWkq/vqNYxwHZEvsXW6dYEo QpODtz2wb5bsyn3Mld9Ucp7jpZNNuRdJP7/wCdeGqwJlzer35nfW/2uAeXW9yoo3 +i2A29EZ/dlYHaJLGsU9RRICgbbR/pGIknsqUei6Ev+AZdPdINppN3/jHS2cn8Bs FWS5gFL5Z+tKIOLsh6EdH6eWieSYzeIv5uVNy4AB9xL7QQO8XKC6s8coiPFODxie Vh+K+Q/vZKO0MSdtRlyfy307YOwix+FdYsrr7FamWkWnZ6AgJqmIOrRwyAFVStDN FfR3CIqm69ZLOaPJ8zavqRh4t1hREMSbBKoWQWJtpPhFl7YDD/PLs0d3F5vZzBQ/ rqk+JMrjpf2JZdnaFs2ynUAmB70VRPzfuw0XNjPXTUyFfdE73WjUxY6H/12g/Azg aQFmh7WADSqz6T8ICn53x6eyOtAReJ4MdTjiqy8eBcxpHy28DEK62jPqm1YzFclY c93pibJUC0r6TA5yBF87xHZMaEOpgMcvsB2QNht6idhA8l9kdJk6yFbDkTIH7TLd dyBLvMY6x0TStmB5AbcEJSzji1O0aZgjMVRpP6m11OliivcuHrbca4K9Bi4prahr V5oOwQkINaw1RdDYI/eQ =7xzi -END PGP SIGNATURE-
[gentoo-user] Fixed bug
Hi all, even if I have been using gentoo for 2 years surely I know to be a newbie again. ;-( Yesterday, upgrading the portage and after an emerge -NDpvu world I received the following massage: Calculating world dependencies / !!! The following installed packages are masked: - app-admin/gnome-system-tools-2.14.0 (masked by: package.mask) /usr/portage/profiles/package.mask: # Gilles Dartiguelongue [EMAIL PROTECTED] (12 Apr 2008) # Masking gnome-system-tools because it is broken, # to help fix it, see bug #214265 - app-admin/system-tools-backends-1.4.2-r1 (masked by: package.mask) \ !!! Ebuilds for the following packages are either all !!! masked or don't exist: app-admin/gnome-system-tools So I went to http://bugs.gentoo.org/show_bug.cgi?id=214265 to fix but I understood absolutely nothing. :-( Could anyone explain to me in which manner I should use the bug page of gentoo? Many thanks emilio
Re: [gentoo-user] e2fsprogs and blocking
On Sun, 2008-08-17 at 08:07 -0400, Chris Walters wrote: The best way to resolve this, I've found is to put the following into your /etc/portage/package.mask file. =sys-libs/e2fsprogs-libs-1.41.0 =sys-fs/e2fsprogs-1.41.0 I masked them because some things depend on com_err, and that is part of e2fsprogs-libs in the 1.41.0 version. Until the upstreams of these packages update their packages, you can't use the new version, unless you're willing to unmerge all of the affected packages. Which is what you should do, as has been mentioned previously in this thread.
Re: [gentoo-user] Fixed bug
On Sun, 2008-08-17 at 14:20 +0200, econti wrote: So I went to http://bugs.gentoo.org/show_bug.cgi?id=214265 to fix but I understood absolutely nothing. :-( Could anyone explain to me in which manner I should use the bug page of gentoo? I really don't understand your question, but I'll explain and if it sounds elementary it's because I don't understand what you're asking. The package is masked because there is a bug associated with it. Unless you are actually capable of fixing the bug (and by your post we'll assume no) then you are powerless until the bug is fixed by someone else. So your only real option at this point is to unmerge gnome-system-tools and possibly follow the bug until if/when it gets fixed. -a
Re: [gentoo-user] Re: Cleaning out my world file
On Sun, 2008-08-17 at 08:00 +0200, Francesco Talamona wrote: It doesn't handle filenames with spaces, I tried with find /etc -xdev -type f -exec qfile -o {} \; that works with spaces, but has awful performances, so that's completely unusable on /usr. BTW I also tried to remove false positive hiding /usr/portage and the overlays folder . Does anyone know how to deal with spaces (to avoid false positives) and keep reasonable running times? Change -print to -print0 and change xargs to xargs -0. I should also say that that script is not fool-proof. It goes under the assumption that files are always going to be installed by the ebuild src_install process. Experience shows this is not always the case. For example the *.pyc files, some, /etc/*, font-cache files, etc. are installed during post-install and so are not recorded by portage as belonging to a package. Common sense always prevails. Of course a person cleaning out an /etc/ dir that you hasn't been cleaned out in five years is going to require considerably more common sense than one who cleaned theirs out last month. I've also modified the script to exclude /usr/portage and /usr/local (they were on separate filesystems for me so I didn't need to). You may need to tweak it further for your needs. -a orphans.sh Description: application/shellscript
Re: [gentoo-user] df and du difference
Quoting Paul Colquhoun [EMAIL PROTECTED]: On Sat, 16 Aug 2008, Ward Poelmans wrote: On Sat, Aug 16, 2008 at 07:50, [EMAIL PROTECTED] wrote: the difference between du and df is about 640 - 188 = 452 MB. and df is showing that my root is full 2.4 times more than du. Normal... Next to the difference due journaling etc, there is one important difference between du en df: Hm. Yeah... Maybe. Journaling add -in my experience- about 32MB. deleted files held open by a running process. du doesn't count these files, df does. Yeap. If there is a file in the /var directory *BEFORE* the /var partition is mounted onto the directory, then du won't find it, but df will know about the space it is using. Yes But you're all missing rounding errors. If you do: du -hcs /path/to/some/dir You'll end up with a different result if you instead do: du -bcs /path/to/some/dir Real example: [EMAIL PROTECTED] ~ $ du -bcs Desktop/ 1289720534 Desktop/ 1289720534 total [EMAIL PROTECTED] ~ $ du -hcs Desktop/ 1.3GDesktop/ 1.3Gtotal If you do the math, 1.3G is more or less 1395864372 bytes; ~110MB in diference and same tool. Regards, Norberto This message was sent using IMP, the Internet Messaging Program.
Re: [gentoo-user] Fixed bug
On Sunday 17 August 2008 08:29:29 Albert Hopkins wrote: On Sun, 2008-08-17 at 14:20 +0200, econti wrote: So I went to http://bugs.gentoo.org/show_bug.cgi?id=214265 to fix but I understood absolutely nothing. :-( Could anyone explain to me in which manner I should use the bug page of gentoo? I really don't understand your question, but I'll explain and if it sounds elementary it's because I don't understand what you're asking. The package is masked because there is a bug associated with it. Unless you are actually capable of fixing the bug (and by your post we'll assume no) then you are powerless until the bug is fixed by someone else. So your only real option at this point is to unmerge gnome-system-tools and possibly follow the bug until if/when it gets fixed. Isn't there also the option of leaving the existing package installed until the bug is fixed? If the package is currently working for him (the bug may be in some section of the program he seldom uses or it may only rear its head in certain configurations or circumstances which don't affect him,) why unmerge what's already there?
Re: [gentoo-user] Fixed bug
On Sun, 2008-08-17 at 11:09 -0400, Daniel D Jones wrote: Isn't there also the option of leaving the existing package installed until the bug is fixed? If the package is currently working for him (the bug may be in some section of the program he seldom uses or it may only rear its head in certain configurations or circumstances which don't affect him,) why unmerge what's already there? That's certainly possible, but I did not perceive that as the OP's question. Admittedly, I didn't quite understand what the OP was asking in the first place.
[gentoo-user] Re: Cleaning out my world file
On Sunday 17 August 2008, Albert Hopkins wrote: On Sun, 2008-08-17 at 08:00 +0200, Francesco Talamona wrote: It doesn't handle filenames with spaces, I tried with find /etc -xdev -type f -exec qfile -o {} \; that works with spaces, but has awful performances, so that's completely unusable on /usr. BTW I also tried to remove false positive hiding /usr/portage and the overlays folder . Does anyone know how to deal with spaces (to avoid false positives) and keep reasonable running times? Change -print to -print0 and change xargs to xargs -0. I should also say that that script is not fool-proof. It goes under the assumption that files are always going to be installed by the ebuild src_install process. Experience shows this is not always the case. For example the *.pyc files, some, /etc/*, font-cache files, etc. are installed during post-install and so are not recorded by portage as belonging to a package. Common sense always prevails. Of course a person cleaning out an /etc/ dir that you hasn't been cleaned out in five years is going to require considerably more common sense than one who cleaned theirs out last month. I've also modified the script to exclude /usr/portage and /usr/local (they were on separate filesystems for me so I didn't need to). You may need to tweak it further for your needs. -a It works very well. It's perfect for the purpose. I already found many leftover here and there. As you say the script gives clues, but obviously it can't be automated. Thanks FT -- Linux Version 2.6.26-gentoo, Compiled #2 PREEMPT Sat Aug 9 20:21:11 CEST 2008 One 1GHz AMD Athlon 64 Processor, 2GB RAM, 2004.04 Bogomips Total aemaeth
[gentoo-user] Rate limiting TCP connections...
I've a Netgear DG834G router - and I connect two machines to it using Ethernet... one Gentoo; one Windows... it works reasonably well... I hit a snag when downloading a large file from Gentoo - for example a multi-meg portage archive. At such times, the Windows PC seems to be given a rather unfair share of the bandwidth... with the Gentoo box getting 60K/s but the windows PC sometimes even failing to establish DNS lookups before they time-out. Is there a straightforward way to make my Gentoo box 'play fair' and not hog 100% of the bandwidth? Does anyone else have this problem?
Re: [gentoo-user] Rate limiting TCP connections...
Quoting [EMAIL PROTECTED]: Is there a straightforward way to make my Gentoo box 'play fair' and not hog 100% of the bandwidth? If your router doesn't give you bandwidth and/or traffic shaping control, you can drop some packets. For example, the following rule will accept up to 50 packets per second and drop the rest. The TCP protocol will retry and slow down. iptables -I INPUT -p tcp -m limit --limit 50/sec -j ACCEPT iptables -I INPUT -p tcp -j DROP Does anyone else have this problem? Yes, everyone using TCP :) You can read Linux Advanced Routing and Traffic Control for more info (http://lartc.org/). Regards, Norberto This message was sent using IMP, the Internet Messaging Program.
Re: [gentoo-user] Rate limiting TCP connections...
Norberto Bensa wrote: Is there a straightforward way to make my Gentoo box 'play fair' and not hog 100% of the bandwidth? If your router doesn't give you bandwidth and/or traffic shaping control, you can drop some packets. For example, the following rule will accept up to 50 packets per second and drop the rest. The TCP protocol will retry and slow down. iptables -I INPUT -p tcp -m limit --limit 50/sec -j ACCEPT iptables -I INPUT -p tcp -j DROP Hmmm - that would likely be rather aggressive - I use the router to shift data between the two PCs at 100mbps - it is only the traffic eventually routed over ADSL which poses a problem. Does anyone else have this problem? Yes, everyone using TCP :) You can read Linux Advanced Routing and Traffic Control for more info (http://lartc.org/). Snappy answer... but I'm pretty sure I've never seen this before - on a wide range of OS and network topologies. I didn't have the same problem with two Windows PCs connected to the same router - and it is always the Gentoo box that wins in the landgrab-war for bandwidth. It might also be worth mentioning that the Gentoo box serves DNS for my LAN - so, the DNS request will get from my Windows PC to my Gentoo box without any problem... so, it is actually a fight between bind on gentoo and the download of packages during emerge that pose my annoyance. I wonder, is it likely relevant that named is running as an ordinary user - while emerge is run as root? I also noticed that incoming emails to my postfix mail server timed out during this period... timeout after RCPT from extern.server.org... It seems odd to me.
Re: [gentoo-user] Rate limiting TCP connections...
On Sunday 17 August 2008 16:32:11 Steve wrote: Norberto Bensa wrote: Is there a straightforward way to make my Gentoo box 'play fair' and not hog 100% of the bandwidth? If your router doesn't give you bandwidth and/or traffic shaping control, you can drop some packets. For example, the following rule will accept up to 50 packets per second and drop the rest. The TCP protocol will retry and slow down. iptables -I INPUT -p tcp -m limit --limit 50/sec -j ACCEPT iptables -I INPUT -p tcp -j DROP Hmmm - that would likely be rather aggressive - I use the router to shift data between the two PCs at 100mbps - it is only the traffic eventually routed over ADSL which poses a problem. You could add a destination parameter with -d but that wouldn't catch any downloads of packages from other than your primary mirror, and I agree that this is not the correct solution anyway. Does anyone else have this problem? Yes, everyone using TCP :) You can read Linux Advanced Routing and Traffic Control for more info (http://lartc.org/). Snappy answer... but I'm pretty sure I've never seen this before - on a wide range of OS and network topologies. I didn't have the same problem with two Windows PCs connected to the same router - and it is always the Gentoo box that wins in the landgrab-war for bandwidth. This should happen. Absent any sort of Quality of Service mechanisms, TCP windowing should adjust so that all flows get a roughly equal share of bandwidth. Certain protocols, primarily p2p, will create multiple flows and so gobble up more than their fair share but even then you shouldn't get starvation to the point where DNS times out. Even if emerge were using bittorrent you shouldn't see that much congestion. It might also be worth mentioning that the Gentoo box serves DNS for my LAN - so, the DNS request will get from my Windows PC to my Gentoo box without any problem... so, it is actually a fight between bind on gentoo and the download of packages during emerge that pose my annoyance. I wonder, is it likely relevant that named is running as an ordinary user - while emerge is run as root? I'm not aware of any mechanism on Linux to set network priority based on the process owner. Linux does support QoS and various types of queues, so it might be possible to, say, set up a priority queue which would have the effect you're seeing. I'm not aware of any way to do queuing based on process owner but Linux is such a powerful and diverse system that no one is familiar with every aspect of it's capabilities. There's lots of stuff you can do that I'm not aware of. If it can be done, it certainly would not be the default setup of any sane package. It would be something you'd have to do yourself. I also noticed that incoming emails to my postfix mail server timed out during this period... timeout after RCPT from extern.server.org... It seems odd to me. It is odd, although unfortunately I'm not sure what to tell you as to how to go about fixing it. I think the first thing I'd do is to make sure that it really is congestion that's causing the problem. You could implement something similar to the iptables rule listed above temporarily and see if that causes the symptoms to change or disappear.
Re: [gentoo-user] Rate limiting TCP connections...
Quoting Steve [EMAIL PROTECTED]: Does anyone else have this problem? Yes, everyone using TCP :) You can read Linux Advanced Routing and Traffic Control for more info (http://lartc.org/). Snappy answer... but I'm pretty sure I've never seen this before - on a wide range of OS and network topologies. I didn't have the same problem with two Windows PCs connected to the same router - Ah!! But Windows (XP) uses TC by default. It doesn't use 20% of the network bandwidth unless you tweak some registry setting and/or disable QoS in network properties. [snip] It seems odd to me. Why? Is pretty obvious what's happening: your Linux box is eating all the bandwidth with the MB download because _by_default_ Linux doesn't do any TC at all. If the iptables thingy was too aggressive, try a --limit-rate (or --rate-limit; I can't never get it right) in wget. Regards, Norberto This message was sent using IMP, the Internet Messaging Program.
[gentoo-user] Roland Puntaier ist außer Haus.
Ich werde ab 18.08.2008 nicht im Büro sein. Ich kehre zurück am 30.08.2008.
[gentoo-user] Re: df and du difference
Alan McKinnon wrote: On Sunday 17 August 2008 01:18:21 Paul Colquhoun wrote: Actually, there is one more way to hide a file from du If there is a file in the /var directory *BEFORE* the /var partition is mounted onto the directory, then du won't find it, but df will know about the space it is using. You will probably need to boot from a live CD of some sort to be able to umount the partitions and check the underlying directory, but it might be worth it there is still space unaccounted for after a reboot. There's a much easier way. As root: mount -o bind / /path/to/some/arb/dir see man mount Thank you very much. That was the problem. some files have been hidden in /mnt/backup. I deleted them and problem is solved. Thanks again Platoali