Justin Pryzby wrote:
> Tom Metro wrote:
>> run-parts probably doesn't involve
>> itself in managing the output of the jobs it runs, and therefore it
>> doesn't know if a job has produced any output.
> Without checking source code, that seems to b
Package: sysklogd
Version: 1.4.1-18
If you modify your /etc/syslog.conf to have a line like:
*.crit @192.168.0.255
restart sysklogd, and then run:
logger -p local0.crit -t foo "message"
you'll observe in /var/log/daemon.log:
Apr 24 04:14:29 host syslogd 1.4.1-18: restart.
Apr 24 04:15:13 host s
Kenneth Pronovici wrote:
It's sometimes difficult to distinguish between output from the
different jobs run as part of cron.daily.
...
It would be great if run-parts could visually distinguish between output
from each job that is run. Possible options include a blank line
between jobs, a ruler
Filippo Giunchedi wrote:
...how about hinting about a debian-specific behavior and pointing the user
to README.Debian from the manpage ?
I've seen that done, where a paragraph is tacked on near the end of man
pages that notes there is Debian-specific behavior and references
README.Debian. Tha
Filippo Giunchedi wrote:
...how to know if the server will correctly reply to alive messages
sent by the client? This also requires knowing what protocol version
is being used.
True. As I recall, that feature requires version 2.0 of the protocol. I
forgot about that.
Of course the ssh client
Package: autossh
Version: 1.4a-3
/usr/share/doc/autossh/README.Debian says:
autossh is wrapped to choose a random port or fall back to 21021 the
real binary is /usr/lib/autossh/autossh
In this version of autossh the commandline switch -M overrides
AUTOSSH_PORT enviroment, as opposed to
Package: autossh
Version: 1.4a-3
/usr/share/doc/autossh/README.Debian says:
[...]
Note: if your ssh client/server version is recent enough you can
obtain similar behaviour by setting ClientAliveCountMax and
ServerAliveInterval in /etc/ssh/sshd_config. See sshd_config(5).
This echoes a rec
Andrew Bartlett <[EMAIL PROTECTED]> wrote:
...Samba's startup code runs before the smb.conf is processed. One
could argue that it is a bug, but I would argue it is also WONTFIX. If
we delay logging until after smb.conf processing, we could not log
incorrect smb.conf statements.
I'd argue th
Package: sysv-rc
Version: 2.86.ds1-38
The following patch causes the intent of the --quiet parameter to
invoke-rc.d to be passed on to the init script, providing the init
script uses /lib/init/vars.sh.
-Tom
diff -u -r1.1 invoke-rc.d
--- invoke-rc.d 2007/07/04 17:02:18 1.1
+++ invoke-rc
Package: logrotate
Version: 3.7.1-3
Thanks for the inclusion of the "dateext" patch, but I've noticed an
inconsistent behavior in the way the subject line is set on mailed logs.
If the log is rotated due to the "rotate" directive, you get a subject
line like:
Subject: /var/log/aptitude-2007
Package: logrotate
Version: 3.7.1-3
Often logs don't include year information, and once a log is emailed,
the original file timestamps are lost.
I propose an enhancement that would grab the file timestamp and
incorporate it into the subject line. Something like:
Subject: /var/log/syslog/mai
Nicolas Kratz wrote:
I have written a patch for logrotate, available at
http://www.janasfam.de/nick/logrotate/
* New "keep" config directive; Overrides "rotate".
(Should close: #42684, #44452, #49681)
* Fixed unnecessary renaming of logs. (Should close: #165548)
It doesn't appear this
Yaroslav Halchenko wrote:
I...got surprised that system wide parameters defined in
/etc/logrotate.conf such as
rotate 4
do not have effect in specific logrotate sections, ie there is no really
default value - it gets assumed to be 1.
This is incorrect. /etc/logrotate.conf does indeed establish
Rudy Gevaert wrote:
/var/log/named/named.log {
[...]
create 644 dnsadm arcs
postrotate
if [ -f /var/run/bind/run/named.pid ]; then
/etc/init.d/bind9 restart > /dev/null
fi
endscript
}
[...]
/etc/cron.daily/logrot
Marc Haber wrote:
logrotate keeps the old versions in a format like basename.1, ...,
basename.9, basename.10, ...
This does not look too pretty in a sorted directory listing. ...
if "rotate 999" is configured, basename.001 should be the
first log file to be created.
It doesn't precisely addres
I'd like to add support to this proposal.
Matt Zimmerman <[EMAIL PROTECTED]> wrote:
It's convention that README.Debian is the place that the user should have
looked first (even if they don't have problems), and it might be good for it
to be more visible.
Absolutely.
In my experience README.De
Package: apt-listchanges
Severity: wishlist
Version: 2.72.5
apt-listchanges is great when installing security updates, but viewing
the change logs is too verbose during distribution upgrades. While it is
always possible to temporarily disable apt-listchanges, that approach
would be less conv
Package: dovecot-imapd
Version: 1.0.rc15-2
Severity: normal
Tags: patch
I found myself needing to run two dovecot-auth servers (one for the IMAP
server and one for deliver), which necessitated running two independent
dovecot master daemons. Rather than create a copy of the init.d script,
I tho
Package: netbase
Severity: wishlist
Attached is a patch to add some common alias names for services in
/etc/services.
Service Alias
---
domain dns
bootps dhcp
microsoft-dssmb
webcachewebproxy
I get the historical reasons why DNS is r
Alexander Koch <[EMAIL PROTECTED]> wrote:
I wonder if it is worth trying to have a /etc/services.local
that automatically gets appended to *any* update of the
/etc/services file the user agrees to.
Well you wouldn't want to literally append to /etc/services, otherwise
you end up with the same
Marco d'Itri wrote:
Tom Metro wrote:
Why do we have two virtual packages serving the same purpose? I'm guessing
one is being phased out and replaced by the other.
Yes. Packages should depend on openbsd-inetd | inet-superserver because
post-etch the dependency from netbase will
Package: atftpd
Version: 0.7-11
Severity: minor
atftpd, when running at its default verbosity, logs a request for a
file, but fails to log an error if the requested file does not exist.
On a client machine:
% tftp -i mythtv.vl.com GET bogus
Error on server : File not found
The relevant bit
Package: tftpd
Version: 0.17-15
Severity: minor
tftpd logs a request for a file, but fails to log an error if the
requested file does not exist.
On a client machine:
% tftp -i mythtv.vl.com GET bogus
Error on server : No such file or directory
The relevant bit logged by tftpd:
Mar 28 23:20
Philippe Wilson wrote:
Although the tftpd package is installed and configured to launch from
inetd, it does not Recommend that an inetd server be installed.
It looks to me like this has been resolved.
According to:
http://packages.debian.org/unstable/net/tftpd
there is a dependency on netbase
Paul Martin wrote:
It seems the cleaner approach would be patching logrotate so that it
calls to an external script to obtain the list of log files...
`/usr/sbin/syslogd-listfiles --weekly` {
weekly
Something like this is a possibility, but there's a lot of work
involved.
OK. Could
Freddie Cash writes:
There's logrotate and various scripts under /etc/cron.daily and
/etc/cron.weekly that also do logrotation (the worst offender is
/etc/cron.*/syslogd which uses savelog).
Why hasn't all this been cleared up and all log rotation done centrally
via logrotate?
See bug:
#32642
please allow log file names to be read from a file, like:
@include /var/cache/syslog/filelist {
This is a variation on my wishlist bug:
#326421 - preprocess config files with m4
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=326421
-Tom
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a
Paul Martin wrote:
They did fix it for the shared scripts, but not for the single file
case.
I've written a patch to fix this, and it will be in 3.7.1-3.
Ah, thanks.
For example, I frequently see emails that look like:
/etc/cron.daily/logrotate:
Null message body; hope that's ok
which
Norbert Nemec writes:
Using a simple setup for logrotate to just rotate everything within a
directory only depending on the file size, I always tripped over a nasty
tweak:
/var/log/syslog/*
{
rotate 100
size 50k
compress
...
works fine most of the time, except that in rare cases th
Package: logrotate
Version: 3.7-5
Suggestion: preprocess config files with m4
Here's the motivation:
I recently asked Martin Schulze, maintainer of the Debian sysklogd
package, why it didn't use logrotate. His explanation was essentially
that logrotate couldn't do everything the sysklogd cron s
This problem still exists in 3.7-5:
-rw-r--r-- 1 root root 8303 Aug 28 13:47 cron.log.2.gz
-rw-r--r-- 1 root root 7677 Aug 28 13:39 cron.log.1
-rw-r--r-- 1 root root 8416 Aug 28 06:33 cron.log.3.gz
which seems strange, as gzip doesn't normally alter the timestamp, as
documented:
gzip
Also...this bug report (Jun 2004):
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=252197
for another issue quotes error output from logrotate version 3.7-1 that
identifies the log being processed when the failure occurred, yet the
version I'm using, 3.7-5, lacks this feature. Was this capabi
Back on 25 Nov 2003 Paul Martin wrote:
I'm hopeful that the next upstream version will help this.
Any news on this bug that dates back to 2001? Is there an upstream bug
report you can post a URL for? Is a patch needed?
-Tom
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of
Package: ntp-server
Version: 1:4.2.0a+stable-2
The cron scripts supplied with ntp-server, such as
/etc/cron.daily/ntp-server, start out with:
if [ -d /var/log/ntpstats ]
...
yet that path may easily change if the user alters the 'statsdir'
setting in /etc/ntp.conf, and the user may be un
34 matches
Mail list logo
