Hi Adrian,
thanks. I just orphaned the package as I don't have the time and
infrastructure (icecast server) anymore to work on this package.
Thanks,
Jochen
Am 16.10.2017 um 10:24 schrieb Adrian Bunk:
On Mon, Oct 16, 2017 at 09:14:53AM +0200, Jochen Friedrich wrote:
Hi Adrian,
you may
Hi Adrian,
you may reupload to DELAYED/0 if you like.
Thanks,
Jochen
Am 14.10.2017 um 11:41 schrieb Adrian Bunk:
Control: tags 873861 + patch
Control: tags 873861 + pending
Dear maintainer,
I've prepared an NMU for darkice (versioned as 1.3-0.2) and uploaded
it to DELAYED/10. Please feel
Hi Guillem,
I was planning on sending the patches to the BTS, and then probably
issue a delayed upload. But maybe you prefer if I push my changes
somewhere and you pull those, and do the the upload yourself? Or for
me to just push to the collab-maint git repo and do a direct upload?
It's OK
Hi Guillem,
cool, thanks.
Regards, Jochen
Am 24.08.2016 um 13:05 schrieb Guillem Jover:
Hi!
On Fri, 2015-03-13 at 20:23:13 +0100, Andreas Beckmann wrote:
Package: snmp-mibs-downloader
Version: 1.1
Severity: serious
User: debian...@lists.debian.org
Usertags: piuparts
a test with piuparts
tags 561578 +wontfix
thanks
As i explained, please take this to debian-legal. As long as the consensus of
the Debian legal Team considers RFCs and MIBs non-free, there's not much i can
do regarding this bug.
Thanks, Jochen
Hi Hector,
as I'm currently unable to work on this package, you might upload in to
DELAYED/0 instead.
Thanks,
Jochen
On 22.02.2012 01:44, Hector Oron wrote:
tags 654285 + patch
tags 654285 + pending
thanks
Dear maintainer,
I've prepared an NMU for isakmpd (versioned as 20041012-7.2) and
Hi Adam,
On Sat, January 22, 2011 05:49, Guillem Jover wrote:
The current version of the server crashes at random times, although
easily reproducibly by doing ranged requests, for example fast
forwarding on a video, or depending on the implementation by
the simple fact that it requests ranges
Hi Lucas,
Restarting network management services: snmpd.
Preparing to replace snmpd 5.4.1~dfsg-12 (using
.../snmpd_5.4.3~dfsg-1_amd64.deb) ...
invoke-rc.d: policy-rc.d denied execution of stop.
Unpacking replacement snmpd ...
invoke-rc.d: policy-rc.d denied execution of start.
Setting up snmpd
Hi Philipp,
But then the version is a release candidate version from 2008, and the popcon
inst count is currently 3. I think it should at least be dropped from squeeze
in order to make clamav supportable.
I will do so in a week if nobody speaks up.
That's fine with me. There are some plans
tags 516801 +pending
thanks
Hi,
I've just committed the update to 5.4.3 to SVN. I plan an upload for next week.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Hi Sandro,
a 0-day NMU for a quite recent bug report is quite rude from the
maintainer POV, in particular because you didn't ping the bug neither
you post a patch on the bug report (as required by devref) and you
didn't upload to a DELAYED queue, giving time to the maintainer to
react. Funny
severity 581007 normal
forcemerge 581007 561578
thanks
Hi Virgil,
Because the package is missing the MIBs, traps that worked perfectly
fine before are now not working. snmptrapd shows these errors at startup:
That's intentional as MIBs can't be redistributed. See #561578 for the details.
Hi Simon,
Package: snmpd
Version: 5.4.2.1~dfsg-5
Severity: grave
since the last upgrade, about any GETNEXT request makes snmpd crash,
first logging an assertion failure, then stumbling over what looks like
a null pointer dereference (address 0x20c).
ii libsnmp15
severity 561124 important
thanks
Hi Bjørn,
I was really surprised to see that it is actually deliberately broken
and has been so for a month, even allowing it to enter testing.
Please try to run e.g.
perl -MSNMP -e 'new SNMP::Session()'
on a system with no MIBs and see the generated
Hi Michael,
It has come to my attention that a lot of maintainers are simply adding
a build-depends on libltdl3-dev to try to solve this problem. This is
not a sufficient solution since your package will still use the
embedded libtool code copy. You need to add '--without-included-ltdl'
to
Hi Petr,
I believe that current dpkg-dev handles architecture wildcards in
debian/control correctly, just try
I'll test this. If this works this is way better than the old method.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org
with a subject of
reassign 498475 snmp-mibs-downloader
thanks
Hi Marco,
Indeed, I work next door to a major internet exchange and downloading
everything with snmp-mibs-downloader was still unacceptably slow.
It is also a very bad practice to distribute a package which will
automatically download files from web
Hi,
the same happens to me on sparc hardware. This is the log i get from the
serial port:
8
Begin: Assembling all MD arrays ... mdadm: No devices listed in conf
file were found.
Failure: failed to assemble all arrays.
Hi Lucas,
During a rebuild of all packages in sid, your package failed to build on
amd64.
Relevant part:
make[1]: Entering directory
`/build/user-libnet-snmp-perl_5.2.0-2-amd64-a5NzFX/libnet-snmp-perl-5.2.0'
PERL_DL_NONLAZY=1 /usr/bin/perl -MExtUtils::Command::MM -e test_harness(0,
Hi Vincent,
I have found the fix in SVN repository for net-snmp package. Net-SNMP
comes with a makefile to download and install MIB. This makefile is now
shipped in place of MIBs.
That's true. Before uploading, I just want to add some installation
dialog asking the user if the MIBS
Hi Kurt,
I was looking at return codes for applications making use of
openssl functions and found this in x509.c:
if (X509_verify(cert, key) == -1) {
log_print(x509_cert_validate: self-signed cert is bad);
return 0;
}
X509_verify returns
Version: 5.4.1~dfsg-12
Steve Langasek schrieb:
But it is precisely against version 5.4.1~dfsg-4 that bug #453123 was
submitted. It's possible that bug #453123 is not the same as bug #391203
with which it has been merged, since 453123 shows a failure to start the
daemon at all rather than a
Hi Mark,
It is marked as fixed in net-snmp/5.4.1~dfsg-1, however it is also marked as
found in net-snmp/5.4.1~dfsg-4 so it still appears in the lenny RC bug list.
From my reading of the bug report, this issue is still effecting lenny, and
Josselin has proposed a fix.
Hi Stephen,
while I was there, I made restart have the same behavior as stop +
start, which seemed like the right thing to do - it's slightly
orthogonal, so feel free to drop that part of the patch if it's not
interesting.
Nope, it's even wrong:
- start shouldn't complain if snmpd is
Hi Filippo,
thanks for the confirmation, is there an upload planned given the trivial
nature
of the fix?
There is one more upload planned (this fix and a language update).
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL
severity 502806 important
thanks
In my opinion, this doesn't qualify as release critical bug as this bug doesn't
case any data loss in a Debian system.
Before applying a path like this, I'd like to have at least some testing and an
audit by an upstream developer.
Thanks,
Jochen
--
To
Hi Josselin,
It has nothing to do with the daemon, but rather, as Ferenc Wagner
diagnosed, with the postinst script not calling db_stop as expected.
Nope, it was a bug with the daemon (snmpd and snmptrapd behaved
differently here) and has been fixed by upstream in 5.4.2. In Debian,
the bug is
forcemerge 391203 453123
thanks
Hi Ferenc,
Of course not leaking file descriptors is a good practice, but it
isn't the responsibility of all the daemons of the world to close all
possible file descriptors their parent might have leaked to them (see
for example
tags 504150 +pending
thanks
Hi Steffen,
Once we get a CVE id for this issue, I'll forward it to this bugreport.
Looks like CVE-2008-4309 has been allocated for this bug.
I added the patch to SVN for the next upload.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a
Marc Lehmann schrieb:
Removing the perl cflags altogether also would ask for trouble as you would
loose the largefile config information.
But that has no relation to the problem at hand.
Fortunately, netsnmp doesn't seem to export data structures with a different
size for largefile
Hi Marc,
As libnetsnmpagent is compiled with embedded perl, the include path is
required to link
against this library.
Uhm, no?
In order to use a library with embedded perl, a program needs to call the
PERL_SYS_INIT3()
and PERL_SYS_TERM() macros. Those macros are defined in the perl
Hi Marc,
As it is, debian makes it impossible to use libsnmp without being forced to
also link against the system perl, which is a limitation which does not exist
with the original package, it's strictly a problem in debian.
Not true. As soon as embedded perl support is enabled (and it's
Hi Marc,
net-snmp-config --cflags outputs this:
-DNETSNMP_ENABLE_IPV6 -g -O2 -g -Wall -O2 -DNETSNMP_USE_INLINE -Ulinux
-Dlinux=linux -D_REENTRANT -D_GNU_SOURCE -DDEBIAN -fno-strict-aliasing -pipe
-I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64
tags 497656 +pending
thanks
Hi Vincent,
Would it be possible to push the fix into lenny? Without it, this
particular binary package is almost unusable.
Hi!
Since the package is mostly unusable, I set the severity to grave. I
will prepare an NMU shortly and upload to
severity 500717 important
thanks
Hi Jan,
Severity: grave
Justification: renders package unusable
Changed to important because the bug does NOT render the package unusable to
everyone.
Since upgrading from etch (stable) to testing, snmpd is reporting this in
syslog:
Sep 30 06:25:22 quark
tags 443420 +pending
thanks
Hi Isaac,
Sorry, I thought I replied to this e-mail last week, but its not
showing up in the bug tracker.
I received your mail and fixed the package in svn. I'll upload a new
version later this week.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
Hi Isaac,
This bug seems to be marked as fixed, but I am having the same problem
described by the bug submitter.
from which to which version are you trying to upgrade?
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL
Hi John,
File /usr/lib/python2.4/site-packages/setuptools/command/build_ext.py,
line 77, in swig_sources
sources = _build_ext.swig_sources(self, sources) or sources
TypeError: swig_sources() takes exactly 3 arguments (2 given)
make: *** [debian/python-build-stamp-2.4] Error 1
sorry,
severity important
tags +upstream
thanks
Hi,
snmpd is running at 100% CPU on 2 of my systems, and will not respond to
polls. I have tried the default config file, which has the same effect.
This is also documented upstream at
Hi Tom,
snmpd is running at 100% CPU on 2 of my systems, and will not respond to
polls. I have tried the default config file, which has the same effect.
I suspect this could be an upstream problem.
Do you have the chance to take 5.3.1-3 (from SID) or 5.4~dfsg-1 (from
experimental) and
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Thomas,
Fine by me. Please submit to http://www.net-snmp.org/patches and I'll take
care
of it.
http://sourceforge.net/tracker/index.php?func=detailaid=1624406group_id=12694atid=312694
nitThe help text should clarify that the default is to
found 370132 5.2.3-6
retitle 370132 snmpd segfaults upon first query [s390]
thanks
I just tested 5.2.3-4 which doesn't expose this bug, cool!
Thanks for the report. However, i leave this bug open until the s390 buildd
is fixed
and a corrected s390 binary has been uploaded.
Thanks to
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Andreas,
Please allow me another remark - I think you should by way force
configure to use sane values. Otherwise, you get errors like in
http://buildd.debian.org/fetch.cgi?pkg=net-snmpver=5.2.3-6arch=powerpcstamp=1166829254file=log
without
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Andreas,
The attached patch makes sure we don't mind about that anymore (though
that is a bad patch, configure should have an option like
--mtab=/etc/mtab or so).
The attached patch adds a configure option --with-mnttab=path. Maybe
it's a good
Hi Thibaut,
I just tested 5.2.3-4 which doesn't expose this bug, cool!
Thanks for the report. However, i leave this bug open until the s390 buildd is
fixed
and a corrected s390 binary has been uploaded.
BTW, /etc/init.d/snmpd restart doesn't do its job: it kills the
daemon without
Hi Steve,
I believe we've had some other misbuilds in the past caused by the s390
buildd not having an /etc/mtab in the build environment. I think this has
been fixed now though, so maybe a rebuild is all that's needed for the
immediate problem?
I'll do yet another upload (replace my patch
Hi Steve,
I believe we've had some other misbuilds in the past caused by the s390
buildd not having an /etc/mtab in the build environment. I think this has
been fixed now though, so maybe a rebuild is all that's needed for the
immediate problem?
Unfortunately, the buildd is not yet fixed.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
Package: snmpd
Version: 5.2.2-3
Severity: important
snmpd no longer works on ia64. It dies when receiving first query.
Here's a sample testcase, running snmpwalk from another box ('dogma'):
Could you retest this with 5.2.3-5? One bug was
Hi Aurelien,
using the libc6-dbg package, I have seen that the getmntent() function
is called with a NULL pointer. This may be due to the fact that on s390
there is in the build log:
checking for mount table location... unknown
whereas on architecture there is:
checking for mount
tags 370132 + help moreinfo
thanks
Hi Aurelien,
snmpd no longer works on ia64. It dies when receiving first query.
I have got the same exact problem on s390. Given that this bug renders
I tested this in a chroot on merulo and raptor by running snmpd as user
and using a different port number,
tag 403079 + confirmed
tag 403079 + pending
clone 403079 -1
reassign -1 autoconf
retitle -1 AC_FUNC_GETMNTENT not defining HAVE_GETMNTENT to 1 but to empty
thanks
Lucas Nussbaum schrieb:
'unreproducible' means that you tried to reproduce the issue, but
couldn't, not just that you _think_ you
tag 403079 + unreproducible
severity 403079 normal
thanks
Hi Lucas,
During a rebuild of all packages in etch, I discovered that your package
failed to build on i386.
ucd-snmp/disk.c:235:36: error: operator '||' has no right operand
ucd-snmp/disk.c:283:36: error: operator '||' has no right
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Lucas,
adduser is not essential, you must depend on it.
Yes, i know. And snmpd already does depend on adduser:
Package: snmpd
Section: net
Priority: optional
Architecture: any
Conflicts: snmp (3.6-3), snmptraplogd
Replaces: snmptraplogd
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Stephen,
It doesn't guarantee it's availability in --purge, though. Just wrap
the calls in
if [ -x /usr/sbin/deluser ]; then
else
echo Can't remove $user
fi
I guess this is the best solution.
Thanks,
Jochen
-BEGIN PGP
Hi Roland,
Maybe a default, fixed set of mibs should be shipped and an additional
directory for additional mibs could be prepared (/var/lib/snmp/mibs)?
IMHO the best solution would be to create the .index file during install
time using an utility. This utility could be reexecuted if a
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
libsnmp-multi-perl not only misses a Build-Dependency on libsnmp-perl, but also
a Dependency.
Without libsnmp-perl installed, libsnmp-multi-perl won't work at all.
Thanks,
Jochen
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (GNU/Linux)
reassign 372599 adduser
thanks
Hi Julien,
while the order of commands in snmpd.postrm might not be optimal,
this is rather a bug of deluser --system returning an error on
non-existing users while the man page states it doesn't.
With adduser 3.85:
# deluser --quiet --system blabla; echo $?
0
Hi Pierre,
uploaded in delayed/2 with urgency set to high.
I just applied the patch mentioned twice on the BTS.
Currently, cheops is the only app still using this ancient SNMP library
(at least on i386, cheops uses libsnmp9 on all other archs and in
Ubuntu, as well). I'd really prefer to
tag 355850 +pending
thanks
Hi Bastian,
| $ objdump -x /usr/lib/libnetsnmp.so | grep NEEDED
| NEEDED libc.so.6
This was caused by a missing aclocal call in the rules. Fixed in SVN.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble?
Hi Steve,
Any word from your sponsor yet? If not, is this fixed source package
available for download somewhere?
The qgis is in unstable since 3 days.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Hi Matthijs,
I've never tested openntpd, but it is the obvious replacement in case
of legal problems with ntp and it has been released with sarge.
I use openntpd and that works better then ntp IMHO.
Last time i checked,
- it doesn't support attached clocks, so no stratum 1
- it only seems
Hi Steve,
The bug log at
http://sourceforge.net/tracker/index.php?func=detailaid=1256697group_id=12694atid=112694
shows that upstream has issued an official update for the soname of
libsnmp, available in patch form at
Hi Roger,
The diff below details the changes required for the rebuild. I intend
to NMU this in two days unless you have any serious problems with the
diff, or you upload a new version before this time. I have uploaded
the build to DELAYED/2-day, but this will be ignored if you upload
within 2
Steve Langasek wrote:
The changelog for net-snmp 5.2.1.2-2 states:
* Added conflicts against all packages using libsnmp5 due to upstream
changing the binary API without increasing SONAME. See: #322500
This is not an appropriate fix for this bug. Conflicts are
contraindicated by policy,
tags 323038 + upstream
forwarded 323038 http://net-snmp.sf.net
thanks
I forwarded the bug upstream - SourceForge #1259049
Thanks for the report.
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
forward 322500 http://net-snmp.sf.net
tag + upstream
thanks
Hi,
unfortunately, simply changing the SONAME would make us incompatible
with the rest of the world. I reported the bug back upstream and hope
they will change the SONAME
upstream.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL
tags 303062 + unreproducible
thanks
Hi Sean,
snip
checking tcpd.h usability... yes
checking tcpd.h presence... yes
checking for tcpd.h... yes
checking for TCP wrappers library -lwrap... no
checking for yp_get_default_domain... no
checking for yp_get_default_domain in -lnsl... yes
Hi Henrique,
Expect an NMU uploaded to 5-DAY in the next 24h if the maintainer does not
reply to this bug. After the upload, he will still have 5 days to remove
the NMU from incoming, or to upload a new version of the package (which
would invalidate the NMU).
I'm away until Saturday, so i
Hi Henrique,
In fact, if you would not mind a temporary workaround in sid, I would very
much like to upload the NMU 0-day ASAP (but I won't do it unless you tell me
to).
Go ahead.
Thanks,
Jochen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact
69 matches
Mail list logo
