Re: [gentoo-user] Emerge difficulty.
Hey Billy, Well, this caused me to look into it more carefully, so I finally umounted the portage directory, and then did an emerge sync to get a new repository on the local machine. Emerge now works. remount portage, doesn't work. You were right, this is an NFS problem. Not sure what type, but I at least know what to look for, and can get my system upgraded in the meantime. I appreciate your input, thanks for your time. I will try and post if I figure out what th eproblem is. I checked the Gentoo wiki, there is a how-to on getting portage on an nfs share, which is what I used to set it up initially. I will have to look and see if I can figure out why it stopped working, which seems bazaar. Michael On Nov 4, 2005, at 10:19 AM, Billy Holmes wrote: Michael A Rowley, MD wrote: works. But whenever I try to get a package to emerge, I get a failure at 'set +x' on the debug screen, and it just sits there. -x The shell shall write to standard error a trace for each command after it expands the command and before it executes it. It is unspecified whether the command that turns tracing off is traced. so 'set +x' disables debugging, and 'set -x' enables it for the shell. Perhaps the script is doing something else right after that... maybe an strace? However, I don't think it's something that complicated. Normally the simplest answer is correct. Check dmesg? have lockd and statd running (since it's an NFS mount)? -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge difficulty.
On Nov 3, 2005, at 1:42 PM, Billy Holmes wrote:Michael A Rowley, MD wrote: + set +x it's possible there is a problem on that volume, and it's stuck waiting for IO? try fscking that volume?Ok, tried this, and it didn't seem to do anything. The portage directory is on a NFS drive on the network, but it works from the other servers on the system, this is the only machine having a problem. I tried doing an emerge regen, and that worked well, also emerge -pv works. But whenever I try to get a package to emerge, I get a failure at 'set +x' on the debug screen, and it just sits there.M -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge difficulty.
Michael A Rowley, MD wrote: works. But whenever I try to get a package to emerge, I get a failure at 'set +x' on the debug screen, and it just sits there. -x The shell shall write to standard error a trace for each command after it expands the command and before it executes it. It is unspecified whether the command that turns tracing off is traced. so 'set +x' disables debugging, and 'set -x' enables it for the shell. Perhaps the script is doing something else right after that... maybe an strace? However, I don't think it's something that complicated. Normally the simplest answer is correct. Check dmesg? have lockd and statd running (since it's an NFS mount)? -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge difficulty.
Michael A Rowley, MD wrote: + set +x it's possible there is a problem on that volume, and it's stuck waiting for IO? try fscking that volume? -- gentoo-user@gentoo.org mailing list
[gentoo-user] Emerge difficulty.
Hello all... I am having a perplexing problem, and I have no idea where to look for an answer. I ahve a server that has been running well for aver a year. I decided to upgrade the SVN server on it, so ran emerge -pv subversion got back: These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild U ] net-www/apache-2.0.54-r31 [2.0.54-r12] +apache2* - debug -doc -ldap -mpm-leader -mpm-peruser -mpm-prefork -mpm- threadpool -mpm-worker -no-suexec (-selinux) +ssl -static-modules - threads 51 kB [ebuild U ] sys-devel/libperl-5.8.6-r1 [5.8.2] +berkdb -debug +gdbm* -ithreads 0 kB [ebuild U ] sys-libs/gdbm-1.8.3-r2 [1.8.0-r5] +berkdb 0 kB [ebuild U ] dev-lang/perl-5.8.6-r6 [5.8.2-r1] +berkdb -build - debug -doc +gdbm* -ithreads -minimal -perlsuid 0 kB [ebuild U ] dev-util/subversion-1.2.3-r2 [1.2.1] +apache2 -bash- completion +berkdb -emacs +java +nls -nowebdav +perl +python +zlib 7,068 kB [ebuild N] app-admin/perl-cleaner-1.01 0 kB Total size of downloads: 7,120 kB then I run emerge -v subversion and get Calculating dependencies ...done! emerge (1 of 6) net-www/apache-2.0.54-r31 to / and then nothing... for ever... let this sit over night, and still nothing... wtf!?!?!? this was working. look in the log shows: 1130962657: Started emerge on: Nov 02, 2005 15:17:37 1130962657: *** emerge --verbose subversion 1130962657: emerge (1 of 6) net-www/apache-2.0.54-r31 to / 1130962657: === (1 of 6) Cleaning (net-www/apache-2.0.54-r31::/usr/ portage/net-www/apache/apache-2.0.54-r31.ebuild) 1130962659: === (1 of 6) Compiling/Merging (net-www/apache-2.0.54- r31::/usr/portage/net-www/apache/apache-2.0.54-r31.ebuild) 1130962682: *** terminating. this was after I terminated the process with a ctl-c. what now? Michael -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge difficulty.
On Wednesday 02 November 2005 20:39, Michael A Rowley, MD wrote: Calculating dependencies ...done! emerge (1 of 6) net-www/apache-2.0.54-r31 to / and then nothing... for ever... let this sit over night, and still nothing... wtf!?!?!? this was working. I've had similar happen to me on a few occasions. The first time when it didn't do anything in quite a while (minute or two) I re-ran the same command with debug (and possible under strace). Emerge would do its stuff then print set +x and hang. I think I eventually tracked it down to waiting on random data. Try copying some data in or out of the machine, preferable over a non-encrypted medium, or some fairly heavy disk i/o (emerge sync, emerge metadata, du -s), if you're logged in locally to the machine bash the keyboard and move the mouse a lot, to generate some entropy while waiting for emerge to do something. -- Mike Williams -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge difficulty.
On Nov 2, 2005, at 4:33 PM, Mike Williams wrote: On Wednesday 02 November 2005 20:39, Michael A Rowley, MD wrote: Calculating dependencies ...done! emerge (1 of 6) net-www/apache-2.0.54-r31 to / and then nothing... for ever... let this sit over night, and still nothing... wtf!?!?!? this was working. I've had similar happen to me on a few occasions. The first time when it didn't do anything in quite a while (minute or two) I re-ran the same command with debug (and possible under strace). Emerge would do its stuff then print set +x and hang. Yup, never thought of doing a --debug... this is what I got.. ...stuff deleted ++ find /var/tmp/portage/apache-2.0.54-r31 -mindepth 1 -maxdepth 1 + '[' -z '' ']' + rmdir /var/tmp/portage/apache-2.0.54-r31 + true + set +x Exiting on signal 2 finally had to break with a ctl-c I think I eventually tracked it down to waiting on random data. Try copying some data in or out of the machine, preferable over a non-encrypted medium, or some fairly heavy disk i/o (emerge sync, emerge metadata, du -s), if you're logged in locally to the machine bash the keyboard and move the mouse a lot, to generate some entropy while waiting for emerge to do something. I am not sure why this would do anything, but id does make for an interesting event. I recently had to pull the monitor off the server to replace a workstation. I usually ssh into the server anyway, so no loss. But can't really sit at the server now. May need to move a monitor back. Why in the hell would doing the above fix the problem? wierd. I need to do some research to see what set +x does. Read the man page, but still kind of scratching my head. Any other ideas? M -- Mike Williams -- gentoo-user@gentoo.org mailing list -- gentoo-user@gentoo.org mailing list
