You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@spamassassin.apache.org by bu...@bugzilla.spamassassin.org on 2012/12/07 17:14:59 UTC
[Bug 6872] New: Net::DNS 0.69 breaks sa-update
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Bug ID: 6872
Summary: Net::DNS 0.69 breaks sa-update
Product: Spamassassin
Version: 3.3.2
Hardware: All
OS: All
Status: NEW
Severity: major
Priority: P2
Component: sa-update
Assignee: dev@spamassassin.apache.org
Reporter: toddr@cpanel.net
Classification: Unclassified
Created attachment 5114
--> https://issues.apache.org/SpamAssassin/attachment.cgi?id=5114&action=edit
Patches sa-update to handle when the returned string is not quoted.
Net::DNS has broken sa-update -
https://rt.cpan.org/Ticket/Display.html?id=81760
It sounds like they're going to fix it but just the same...
This patch to sa-update would work around Net::DNS 0.70. $1 should never be
used anyways unless you're sure the regex matched. It's just asking for
trouble. Also local $1 makes me nervous. It could potentially behave different
on different versions of perl depending on if bugs existed with local or $1.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Todd Rinaldo <to...@cpanel.net> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |toddr@cpanel.net
--- Comment #1 from Todd Rinaldo <to...@cpanel.net> ---
Sorry, I missed the mailing list replies on this. Feel free to close.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #16 from Mark Martinec <Ma...@ijs.si> ---
> Good catch, and sorry about the duplicate bug report.
>
> On an informative note, which I will send to the perl ports manager as well;
> the current FreeBSD userbase is in for a treat...
>
> [root@firefly:~](1)>head /usr/ports/dns/p5-Net-DNS/Makefile
> # Created by: James FitzGibbon <jf...@FreeBSD.org>
> # $FreeBSD: ports/dns/p5-Net-DNS/Makefile,v 1.90 2012/12/17 05:50:12 svnexp
> Exp $
> PORTNAME= Net-DNS
> PORTVERSION= 0.70
The port is at 0.71 since 2012-12-18, you have a stale copy:
# Created by: James FitzGibbon <jf...@FreeBSD.org>
# $FreeBSD: ports/dns/p5-Net-DNS/Makefile,v 1.91 2012/12/18 09:25:02 svnexp Exp
$
PORTNAME= Net-DNS
PORTVERSION= 0.71
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Mark Martinec <Ma...@ijs.si> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target Milestone|Undefined |3.4.0
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #13 from Mark Martinec <Ma...@ijs.si> ---
Created attachment 5124
--> https://issues.apache.org/SpamAssassin/attachment.cgi?id=5124&action=edit
the change in r1425068, replaces $rr->rdatastr with $rr->txtdata where
appropriate
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #8 from Mark Martinec <Ma...@ijs.si> ---
(In reply to comment #3, by Willem Toorop)
Now back to the original problem.
> I believe you can also work around the issue by putting a space in the
> version TXT resource record and an @ in the mirror TXT resource record,
> like this:
>
> 2.3.3.updates.spamassassin.org. TXT "1418219 "
> mirrors.updates.spamassassin.org. TXT
> "http://anonymous@spamassassin.apache.org/updates/MIRRORED.BY"
>
> The space and the @ will force a double quoted return value from rdatastr
> even with the new Net::DNS.
> I have tested this with sa-update version svn917659 running on Perl version
> 5.12.4 on FreeBSD ...
This looks like a good and innocent-enough workaround for
a bug in SpamAssassin caused by the change in Net::DNS.
I tested it with SA 3.3 (uses wget) and 3.4(trunk) (uses curl or
wget or fetch or wget), and it seems to do the job, while allowing
Net::DNS folks to proceed with their wish of retaining this new behaviour.
(there is a lot to be said about the incorrect assumption throughout
most of the SpamAssassin code that the Net::DNS API is 8-bit clean,
conflicting with their (silly, in my view) concept that the API uses
the same encoding as zone ascii text files - but I'll leave that to
another posting)
> I would like to know if your going to release with the fix (or apply the
> trick just mentioned) shortly, to asses whether we need to revert the
> behaviour of TXT RR's or not...
Considering the time it takes/took some distributions and some installations
to pick up a new version of SpamAssassin in the past, I think it is futile
to think of going into trouble of releasing a 3.3.3 just to fix this problem,
knowing that it will only reach a very limited audience.
I think adopting Willem's suggestion is a good (yet hackish) short term
solution until the 3.4 can be released to a general public (noting that
several sites are already using 3.4(trunk) in production).
> PS. I have the intention to e-mail about upcoming releases to developers
> that use Net::DNS in their packages in the future, hopefully with the effect
> that such nasty issues come up before release.
Thanks, good idea.
> PS2. I have tested new Net::DNS with spamassassin before release, but did
> not notice this issue because sa-update was quiet as normal.
Neither did I :)
> Shouldn't sa-update at least give a warning when it is unable to determine
> whether it needs to update or not?
It does, as far as I can tell:
spamassassin-3.3(branch):
# sa-update
http: GET 3.3 request failed: 400 URL must be absolute: 400 URL must be
absolute
error: no mirror data available for channel updates.spamassassin.org
channel: MIRRORED.BY contents were missing, channel failed
(exit status 4)
spamassassin(trunk):
# sa-update
channel: no 'mirrors.updates.spamassassin.org' record found, channel failed
(exit status 4)
and with "1418219 " but without "anonymous@":
# sa-update
error: no mirror data available for channel updates.spamassassin.org
channel: MIRRORED.BY file location was not in DNS, channel failed
(exit status 4)
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Willem Toorop <wi...@nlnetlabs.nl> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |willem@nlnetlabs.nl
--- Comment #3 from Willem Toorop <wi...@nlnetlabs.nl> ---
Hi,
I'm responsible for the new Net::DNS release.
I believe you can also work around the issue by putting a space in the version
TXT resource record and an @ in the mirror TXT resource record, like this:
2.3.3.updates.spamassassin.org. TXT "1418219 "
mirrors.updates.spamassassin.org. TXT
"http://anonymous@spamassassin.apache.org/updates/MIRRORED.BY"
The space and the @ will force a double quoted return value from rdatastr even
with the new Net::DNS.
I have tested this with sa-update version svn917659 running on Perl version
5.12.4 on FreeBSD by putting those records in updates.spamassassin.net-dns.org
and changing sa-update with:
my @channels = ( 'updates.spamassassin.net-dns.org' );
I suppose you can judge best for yourself if this would work on other versions
of sa-update and if it is a "proper" workaround.
I would like to know if your going to release with the fix (or apply the trick
just mentioned) shortly, to asses whether we need to revert the behaviour of
TXT RR's or not...
Best regards,
Willem Toorop
PS. I have the intention to e-mail about upcoming releases to developers that
use Net::DNS in their packages in the future, hopefully with the effect that
such nasty issues come up before release.
PS2. I have tested new Net::DNS with spamassassin before release, but did not
notice this issue because sa-update was quiet as normal. Shouldn't sa-update at
least give a warning when it is unable to determine whether it needs to update
or not?
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #15 from tycho.eggen@gmail.com ---
Good catch, and sorry about the duplicate bug report.
On an informative note, which I will send to the perl ports manager as well;
the current FreeBSD userbase is in for a treat...
========================================================================(215)[root@firefly:~](1)>head
/usr/ports/dns/p5-Net-DNS/Makefile
# Created by: James FitzGibbon <jf...@FreeBSD.org>
# $FreeBSD: ports/dns/p5-Net-DNS/Makefile,v 1.90 2012/12/17 05:50:12 svnexp Exp
$
PORTNAME= Net-DNS
PORTVERSION= 0.70
CATEGORIES= dns net perl5 ipv6
MASTER_SITES= CPAN
PKGNAMEPREFIX= p5-
MAINTAINER= perl@FreeBSD.org
========================================================================
I ran into this issue while doing a fresh install.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Mark Martinec <Ma...@ijs.si> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |tycho.eggen@gmail.com
--- Comment #14 from Mark Martinec <Ma...@ijs.si> ---
*** Bug 6882 has been marked as a duplicate of this bug. ***
--
You are receiving this mail because:
You are the assignee for the bug.
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, Mark Martinec wrote:
>> Please wait a little before testing, let me make sure everything is
>> alright.
>
> Ok, go now... (another commit)
Looks like that did it.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The fetters imposed on liberty at home have ever been forged out
of the weapons provided for defense against real, pretended, or
imaginary dangers from abroad. -- James Madison, 1799
-----------------------------------------------------------------------
3 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/12/2012 1:42 PM, Mark Martinec wrote:
>> Please wait a little before testing, let me make sure everything is
>> alright.
> Ok, go now... (another commit)
>
> Mark
Hmm, I thought masscheck had a mechanism for which version of the code
so that all masscheckers use the same code. Or is that something I
thought would be a good idea...
Regads,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
> Please wait a little before testing, let me make sure everything is
> alright.
Ok, go now... (another commit)
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/12/2012 09:39 PM, Axb wrote:
>> Great job, folks!
ACK! works great when run correctly <blush>
I'm seeing LOTS of "scantime=1.2" which is perfect!
Thanks Marc for the fast fix!
Axb
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/12/2012 07:48 PM, Marc Andre Selig wrote:
> On Wed, Dec 12, 2012 at 07:38:27PM +0100, Marc Andre Selig wrote:
>
>> Here's the histogram of scan times so far:
>
> And here's the histogram of the first few messages with svn revision 1420907:
>
> mas@falun:~$ cat ~sacheck/masscheckwork/spamassassin-trunk/masses/*.log | perl -ne '/scantime=([0-9]+),/ && $h[$1]++; END{foreach $i (1..20) {printf "%2d: %d\n", $i, $h[$i]}}'
> 1: 72
> 2: 16
> 3: 1
> 4: 4
> 5: 31
> 6: 22
> 7: 9
> 8: 1
> 9: 6
> 10: 3
> 11: 0
> 12: 0
> 13: 0
> 14: 0
> 15: 0
> 16: 0
> 17: 0
> 18: 0
> 19: 0
> 20: 0
>
> Great job, folks!
>
I must be doing something wrong... still seein 15 sec scan times
back to the drawing board
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, Marc Andre Selig wrote:
> And here's the histogram of the first few messages with svn revision 1420907:
>
> mas@falun:~$ cat ~sacheck/masscheckwork/spamassassin-trunk/masses/*.log | perl -ne '/scantime=([0-9]+),/ && $h[$1]++; END{foreach $i (1..20) {printf "%2d: %d\n", $i, $h[$i]}}'
> 1: 72
> 2: 16
> 3: 1
> 4: 4
> 5: 31
> 6: 22
> 7: 9
> 8: 1
> 9: 6
> 10: 3
> 11: 0
> 12: 0
> 13: 0
> 14: 0
> 15: 0
> 16: 0
> 17: 0
> 18: 0
> 19: 0
> 20: 0
>
> Great job, folks!
>
> Regards
> Marc
>
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The social contract exists so that everyone doesn't have to squat
in the dust holding a spear to protect his woman and his meat all
day every day. It does not exist so that the government can take
your spear, your meat, and your woman because it knows better what
to do with them. -- Dagny @ Ace of Spades
-----------------------------------------------------------------------
3 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Marc Andre Selig <a2...@sedacon.com>.
On Wed, Dec 12, 2012 at 07:38:27PM +0100, Marc Andre Selig wrote:
> Here's the histogram of scan times so far:
And here's the histogram of the first few messages with svn revision 1420907:
mas@falun:~$ cat ~sacheck/masscheckwork/spamassassin-trunk/masses/*.log | perl -ne '/scantime=([0-9]+),/ && $h[$1]++; END{foreach $i (1..20) {printf "%2d: %d\n", $i, $h[$i]}}'
1: 72
2: 16
3: 1
4: 4
5: 31
6: 22
7: 9
8: 1
9: 6
10: 3
11: 0
12: 0
13: 0
14: 0
15: 0
16: 0
17: 0
18: 0
19: 0
20: 0
Great job, folks!
Regards
Marc
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Marc Andre Selig <a2...@sedacon.com>.
On Wed, Dec 12, 2012 at 07:28:00PM +0100, Axb wrote:
> Still not good:
[...]
> 15 sec - still way too long
> can anybody reproduce?
Yup. I aborted after a few minutes because it appeared too slow, but
did not post because you had already volunteered.
Here's the histogram of scan times so far:
mas@falun:~$ cat ~sacheck/masscheckwork/weekly_mass_check/masses/*.log | perl -ne '/scantime=([0-9]+),/ && $h[$1]++; END{foreach $i (1..20) {printf "%2d: %d\n", $i, $h[$i]}}'
1: 0
2: 0
3: 0
4: 0
5: 0
6: 0
7: 0
8: 0
9: 0
10: 0
11: 1
12: 5
13: 9
14: 20
15: 24
16: 8
17: 1
18: 1
19: 1
20: 0
Regards
Marc
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Marc Andre Selig <a2...@sedacon.com>.
On Wed, Dec 12, 2012 at 10:37:41AM -0800, John Hardin wrote:
> Just checking: that does involve "svn up" and a make?
On my part, just a re-run of auto-mass-check.sh, which is supposed to
rsync "tagged_builds/weekly_mass_check". If nothing has been tagged yet,
please disregard my previous message. ;)
Regards
Marc
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, Axb wrote:
> On 12/12/2012 07:18 PM, Axb wrote:
>> On 12/12/2012 07:16 PM, John Hardin wrote:
>> > On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
>> >
>> > > Bug 6872, fix breakage introduced by my previous change - DNS replies
>> > > with no answers would wait till their timeout
>> >
>> > Ouch.
>> >
>> > Yeah, that would explain the masscheck behavior. Anybody want to fire
>> > off a test and see if this fixed the long runtimes?
>> >
>>
>> Will do asap - left my test config in place last night...
>
> Still not good:
>
> 15 sec - still way too long
> can anybody reproduce?
Just checking: that does involve "svn up" and a make?
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The social contract exists so that everyone doesn't have to squat
in the dust holding a spear to protect his woman and his meat all
day every day. It does not exist so that the government can take
your spear, your meat, and your woman because it knows better what
to do with them. -- Dagny @ Ace of Spades
-----------------------------------------------------------------------
3 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/12/2012 07:18 PM, Axb wrote:
> On 12/12/2012 07:16 PM, John Hardin wrote:
>> On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
>>
>>> Bug 6872, fix breakage introduced by my previous change - DNS replies
>>> with no answers would wait till their timeout
>>
>> Ouch.
>>
>> Yeah, that would explain the masscheck behavior. Anybody want to fire
>> off a test and see if this fixed the long runtimes?
>>
>
> Will do asap - left my test config in place last night...
Still not good:
Y 6 /home/foo.woas/Maildir/cur//1355297773.28901_2.inet4:2,
AXB_XMAILER_MIMEOLE_OL_4379D,HTML_MESSAGE,KB_RATWARE_MSGID,PYZOR_CHECK,T_BIG_HEADERS_3K,T_FSL_HELO_NON_FQDN_2,T_HELO_MISC_IP,T_KHOP_BIG_TO_CC,T_LONG_HEADER_LINE_80,T_MSGID_NOFQDN2,UNPARSEABLE_RELAY,__ANY_OUTLOOK_MUA,__ANY_TEXT_ATTACH,__ANY_TEXT_ATTACH_DOC,__AXB_MO_OL_03F7E,__AXB_MO_OL_1CC6A,__AXB_MO_OL_37CCA,__AXB_MO_OL_4379D,__AXB_MO_OL_5BD7E,__AXB_MO_OL_6B767,__AXB_MO_OL_8B970,__AXB_MO_OL_AE569,__AXB_MO_OL_C9F34,__AXB_MO_OL_EF682,__AXB_XM_OL_285BC,__AXB_XM_OL_4379D,__BOUNCE_RPATH_NULL,__CT,__CTYPE_HAS_BOUNDARY,__CTYPE_MULTIPART_ALT,__CTYPE_MULTIPART_ANY,__DOS_RCVD_WED,__DOS_RELAYED_EXT,__FH_HAS_XMSMAIL,__FH_HAS_XPRIORITY,__FROM_ENCODED_B64,__FROM_FULL_NAME,__FROM_NEEDS_MIME,__HAS_CC,__HAS_DATE,__HAS_FROM,__HAS_MESSAGE_ID,__HAS_MIMEOLE,__HAS_MSGID,__HAS_MSMAIL_PRI,__HAS_RCVD,__HAS_SUBJECT,__HAS_TO,__HAS_XMAIL,__HAS_X_MAILER,__HAVE_BOUNCE_RELAYS,__HELO_MISC_IP,__HIGHBITS,__KB_MSGID_OUTLOOK_888,__LAST_EXTERNAL_RELAY_NO_AUTH,__LAST_UNTRUSTED_RELAY_NO_AUTH,__MANY_RECIPS,__MIMEOLE_MS,__M
I
ME_HTML,__MIME_QP,__MIME_VERSION,__MISSING_REF,__MISSING_REPLY,__MISSING_THREAD,__MSGID_DOLLARS_MAYBE,__MSGID_DOLLARS_OK,__MSGID_OK_HEX,__MSGID_RANDY,__NONEMPTY_BODY,__NO_INR_YES_REF,__OE_MSGID_2,__OE_MUA,__RCVD_IN_ZEN,__RCVD_VIA_APNIC_E,__RCVD_VIA_RIPE,__RCVD_VIA_RIPE_E,__SANE_MSGID,__SUBJECT_ENCODED_B64,__SUBJECT_NEEDS_MIME,__SUBJ_NOT_SHORT,__TAG_EXISTS_BODY,__TAG_EXISTS_HEAD,__TAG_EXISTS_HTML,__TAG_EXISTS_META,__TAG_EXISTS_STYLE,__TOCC_EXISTS,__TVD_BODY,__TVD_MIME_ATT_TP,__UNPARSEABLE_RELAY_COUNT,__XM_MSOE6,__XM_MS_IN_GENERAL,__XM_OUTLOOK_EXPRESS,__XPRIO
time=1355297773,scantime=15,format=f,reuse=no,set=1
15 sec - still way too long
can anybody reproduce?
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/12/2012 07:16 PM, John Hardin wrote:
> On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
>
>> Bug 6872, fix breakage introduced by my previous change - DNS replies
>> with no answers would wait till their timeout
>
> Ouch.
>
> Yeah, that would explain the masscheck behavior. Anybody want to fire
> off a test and see if this fixed the long runtimes?
>
Will do asap - left my test config in place last night...
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by da...@chaosreigns.com.
On 12/12, John Hardin wrote:
> Yeah, that would explain the masscheck behavior. Anybody want to
> fire off a test and see if this fixed the long runtimes?
Nice, thanks. It seems like we should make some attempt at monitoring for
drastic changes in masscheck run time. Like, I don't know, keep a log of
how long it takes each day, and maybe email an error to the person running
it if it runs a lot longer than usual.
--
"The whole aim of practical politics is to keep the populace alarmed --
and hence clamorous to be led to safety -- by menacing it with an endless
series of hobgoblins, all of them imaginary." - H. L. Mencken
http://www.ChaosReigns.com
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/12/2012 07:30 PM, Mark Martinec wrote:
>>> Bug 6872, fix breakage introduced by my previous change - DNS replies
>>> with no answers would wait till their timeout
>>
>> Ouch.
>>
>> Yeah, that would explain the masscheck behavior. Anybody want to fire off
>> a test and see if this fixed the long runtimes?
>
> Please wait a little before testing, let me make sure everything is alright.
ooops - ok - just ran a test and it was still taking 15 sec.
shoot me a direct mail when you need a test run.
Axb
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
> > Bug 6872, fix breakage introduced by my previous change - DNS replies
> > with no answers would wait till their timeout
>
> Ouch.
>
> Yeah, that would explain the masscheck behavior. Anybody want to fire off
> a test and see if this fixed the long runtimes?
Please wait a little before testing, let me make sure everything is alright.
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
> Bug 6872, fix breakage introduced by my previous change - DNS replies
> with no answers would wait till their timeout
Ouch.
Yeah, that would explain the masscheck behavior. Anybody want to fire off
a test and see if this fixed the long runtimes?
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The social contract exists so that everyone doesn't have to squat
in the dust holding a spear to protect his woman and his meat all
day every day. It does not exist so that the government can take
your spear, your meat, and your woman because it knows better what
to do with them. -- Dagny @ Ace of Spades
-----------------------------------------------------------------------
3 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
> Bug 6872, fix breakage introduced by my previous change - DNS replies
> with no answers would wait till their timeout
Ouch.
Yeah, that would explain the masscheck behavior. Anybody want to fire off
a test and see if this fixed the long runtimes?
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The social contract exists so that everyone doesn't have to squat
in the dust holding a spear to protect his woman and his meat all
day every day. It does not exist so that the government can take
your spear, your meat, and your woman because it knows better what
to do with them. -- Dagny @ Ace of Spades
-----------------------------------------------------------------------
3 days until Bill of Rights day
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #5 from Mark Martinec <Ma...@ijs.si> ---
Created attachment 5117
--> https://issues.apache.org/SpamAssassin/attachment.cgi?id=5117&action=edit
fix my previous patch
Bug 6872, fix breakage introduced by my previous change - DNS replies
with no answers would wait till their timeout
Sending lib/Mail/SpamAssassin/DnsResolver.pm
Committed revision 1420890.
--
You are receiving this mail because:
You are the assignee for the bug.
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Axb <ax...@gmail.com>.
On 12/13/2012 08:59 PM, darxus@chaosreigns.com wrote:
> On 12/13, Kevin A. McGrail wrote:
>> I think this is very good. We need to get masscheck unified so
>> everyone is using the same version. Alex, did you ever get the
>> minimal version checked in so perhaps we can start building on that
>> version as the official "protect" version?
>
> Yes, and I updated the nightly masscheck wiki page to point to it.
>
Yes, and Darxus kindly updated the nightly masscheck wiki page to point
to it.
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Thu, 13 Dec 2012, darxus@chaosreigns.com wrote:
> On 12/13, John Hardin wrote:
>> On Thu, 13 Dec 2012, darxus@chaosreigns.com wrote:
>>
>>> On 12/13, Kevin A. McGrail wrote:
>>>> I think this is very good. We need to get masscheck unified so
>>>> everyone is using the same version. Alex, did you ever get the
>>>> minimal version checked in so perhaps we can start building on that
>>>> version as the official "protect" version?
>>>
>>> Yes, and I updated the nightly masscheck wiki page to point to it.
>>
>> Which page is that? The latest change the wiki is reporting occurred
>> on 12/11 and it wasn't related to the masscheck process...
>>
>> http://wiki.apache.org/spamassassin/NightlyMassCheck was last
>> updated in October.
>
> And the description of that update in October is "Update from externally
> hosted auto-mass-check script to automasscheck-minimal in trunk"
Ah, ok. Sorry for the misunderstanding.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
Perfect Security and Absolute Safety are unattainable; beware
those who would try to sell them to you, regardless of the cost,
for they are trying to sell you your own slavery.
-----------------------------------------------------------------------
2 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by da...@chaosreigns.com.
On 12/13, John Hardin wrote:
> On Thu, 13 Dec 2012, darxus@chaosreigns.com wrote:
>
> >On 12/13, Kevin A. McGrail wrote:
> >>I think this is very good. We need to get masscheck unified so
> >>everyone is using the same version. Alex, did you ever get the
> >>minimal version checked in so perhaps we can start building on that
> >>version as the official "protect" version?
> >
> >Yes, and I updated the nightly masscheck wiki page to point to it.
>
> Which page is that? The latest change the wiki is reporting occurred
> on 12/11 and it wasn't related to the masscheck process...
>
> http://wiki.apache.org/spamassassin/NightlyMassCheck was last
> updated in October.
And the description of that update in October is "Update from externally
hosted auto-mass-check script to automasscheck-minimal in trunk"
--
Immorality: "The morality of those who are having a better time"
- Henry Louis Mencken
http://www.ChaosReigns.com
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Thu, 13 Dec 2012, darxus@chaosreigns.com wrote:
> On 12/13, Kevin A. McGrail wrote:
>> I think this is very good. We need to get masscheck unified so
>> everyone is using the same version. Alex, did you ever get the
>> minimal version checked in so perhaps we can start building on that
>> version as the official "protect" version?
>
> Yes, and I updated the nightly masscheck wiki page to point to it.
Which page is that? The latest change the wiki is reporting occurred on
12/11 and it wasn't related to the masscheck process...
http://wiki.apache.org/spamassassin/NightlyMassCheck was last updated in
October.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The Tea Party wants to remove the Crony from Crony Capitalism.
OWS wants to remove Capitalism from Crony Capitalism.
-- Astaghfirullah
-----------------------------------------------------------------------
2 days until Bill of Rights day
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by da...@chaosreigns.com.
On 12/13, Kevin A. McGrail wrote:
> I think this is very good. We need to get masscheck unified so
> everyone is using the same version. Alex, did you ever get the
> minimal version checked in so perhaps we can start building on that
> version as the official "protect" version?
Yes, and I updated the nightly masscheck wiki page to point to it.
--
"It's a dangerous business, Frodo, going out your front door. You step
into the Road, and if you don't keep your feet, there is no knowing
where you might be swept off to." - Bilbo Baggins
http://www.ChaosReigns.com
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/13/2012 1:59 AM, Marc Andre Selig wrote:
> On Wed, Dec 12, 2012 at 10:45:59AM -0800, John Hardin wrote:
>
>> If it still misbehaves, I'd suggest a manual "svn up ; make" just to
>> see if the masscheck scripting isn't updating the code properly for
>> some reason, which would be a different problem.
> masscheck scripting does not use subversion (directly); in fact, issuing
> "svn update" from the masscheck work directory does nothing since the
> subversion directories are not present. Instead, auto-mass-check.sh uses
> rsync to download a tagged version. Also, I believe make is superfluous
> for masscheck users as the only thing it builds is the perceptron,
> which is not called from auto-mass-check.sh.
>
> Attached are a diff and a modified version of auto-mass-check.sh to
> add two new command line options "--svn-weekly" and "--svn-nightly" to
> check out or update to svn head and execute from there. To minimize
> unexpected side effects, it does not upload results when run with
> one of these options. (It also includes an enhancement to enable the
> HitFreqsRuleTiming plugin if a variable FREQSTIMING is set to a nonzero
> string in .auto-mass-check.cf.)
>
>
I think this is very good. We need to get masscheck unified so everyone
is using the same version. Alex, did you ever get the minimal version
checked in so perhaps we can start building on that version as the
official "protect" version?
regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Marc Andre Selig <a2...@sedacon.com>.
On Wed, Dec 12, 2012 at 10:45:59AM -0800, John Hardin wrote:
> If it still misbehaves, I'd suggest a manual "svn up ; make" just to
> see if the masscheck scripting isn't updating the code properly for
> some reason, which would be a different problem.
masscheck scripting does not use subversion (directly); in fact, issuing
"svn update" from the masscheck work directory does nothing since the
subversion directories are not present. Instead, auto-mass-check.sh uses
rsync to download a tagged version. Also, I believe make is superfluous
for masscheck users as the only thing it builds is the perceptron,
which is not called from auto-mass-check.sh.
Attached are a diff and a modified version of auto-mass-check.sh to
add two new command line options "--svn-weekly" and "--svn-nightly" to
check out or update to svn head and execute from there. To minimize
unexpected side effects, it does not upload results when run with
one of these options. (It also includes an enhancement to enable the
HitFreqsRuleTiming plugin if a variable FREQSTIMING is set to a nonzero
string in .auto-mass-check.cf.)
Hope this helps.
Regards
Marc
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by John Hardin <jh...@impsec.org>.
On Wed, 12 Dec 2012, bugzilla-daemon@issues.apache.org wrote:
> https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
>
> --- Comment #7 from Mark Martinec <Ma...@ijs.si> ---
> Still not quite right, this should be better:
>
> Sending lib/Mail/SpamAssassin/DnsResolver.pm
> Committed revision 1420906.
Another change checked in. Please try again.
If it still misbehaves, I'd suggest a manual "svn up ; make" just to see
if the masscheck scripting isn't updating the code properly for some
reason, which would be a different problem.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
The social contract exists so that everyone doesn't have to squat
in the dust holding a spear to protect his woman and his meat all
day every day. It does not exist so that the government can take
your spear, your meat, and your woman because it knows better what
to do with them. -- Dagny @ Ace of Spades
-----------------------------------------------------------------------
3 days until Bill of Rights day
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #7 from Mark Martinec <Ma...@ijs.si> ---
Still not quite right, this should be better:
Sending lib/Mail/SpamAssassin/DnsResolver.pm
Committed revision 1420906.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Mark Martinec <Ma...@ijs.si> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |L@rs.Timmann.de
--- Comment #11 from Mark Martinec <Ma...@ijs.si> ---
*** Bug 6878 has been marked as a duplicate of this bug. ***
--
You are receiving this mail because:
You are the assignee for the bug.
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
> On 12/14/2012 12:33 PM, Mark Martinec wrote:
> > 2) ask the Net::DNS folks if they'd be willing to put back the
> >
> > unconditionally double-quoted form for 0.71, and keep it there
> > for at least until SpamAssassin 3.4 really hits the streets;
> > meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
> >
> > ...
> > Looking back at what I just wrote, the #2 looks to me
> > the most appealing.
>
> +1. Can you reach out to them and just ask for perhaps 3 months?
After an offline conversation with Willem Toorop over the weekend,
they have now released 0.71 with several bug fixes, along with
temporarily reverting the double quoting issue, until we get
our release fixed.
This is from the Net::DNS 0.71 'Changes' file:
**** 0.71 Dec 15, 2012
Temporary workaround rt.cpan.org #81760
The rdatastr method for TXT RRs will return unconditionally
quoted rdata fields to work around an issue with updating
SpamAssassin rules. This workaround will be reverted after
release of a version of SpamAssassin which resolves the issue.
Meanwhile we should be suggesting people to either apply the
sa-update fix, or go for Net::DNS 0.71, or stick with 0.68,
but stay away froom Net::DNS 0.69 and 0.70.
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
Willem,
> Hi Spamassassin developers,
>
> This message was from more than a year ago. We would like to revert our
> temporary workaround in Net::DNS. Is it safe now? Is a Spamassassin
> released which works without the workaround?
>
> op 14-12-12 18:37, Kevin A. McGrail schreef:
> > On 12/14/2012 12:33 PM, Mark Martinec wrote:
> >> 2) ask the Net::DNS folks if they'd be willing to put back the
> >>
> >> unconditionally double-quoted form for 0.71, and keep it there
> >> for at least until SpamAssassin 3.4 really hits the streets;
> >> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
> >>
> >> ...
> >> Looking back at what I just wrote, the #2 looks to me
> >> the most appealing.
> >
> > +1. Can you reach out to them and just ask for perhaps 3 months?
> >
> > Regards,
> > KAM
Well, the three months grace period has extended itself to a year,
and we are stuck at 3.4.0-RC3 since October. I don't think there are
any major blockers for a release, just needs some good will on our part
to do it.
So I don't have a heart to ask you to stall the pending change in Net::DNS.
On the bright side, of those that keep up with most recent releases of
perl modules, many of us are already running the unreleased 3.4.0 for a
long time (including some derivatives like zimbra), so won't be affected.
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Willem Toorop <wi...@nlnetlabs.nl>.
Hi Kevin,
No problem. We will wait until a new spamassassin is released.
Regards,
-- Willem
op 07-01-14 22:13, Kevin A. McGrail schreef:
> On 12/24/2013 7:34 AM, Willem Toorop wrote:
>> Hi Spamassassin developers,
>>
>> This message was from more than a year ago. We would like to revert our
>> temporary workaround in Net::DNS. Is it safe now? Is a Spamassassin
>> released which works without the workaround?
>>
>> Best regards,
>>
>> -- Willem
> Willem,
>
> Could you give us one more month please? You are right it has been far
> too long and I'm working on this now.
>
> Regards,
> KAM
>>
>> op 14-12-12 18:37, Kevin A. McGrail schreef:
>>> On 12/14/2012 12:33 PM, Mark Martinec wrote:
>>>> 2) ask the Net::DNS folks if they'd be willing to put back the
>>>> unconditionally double-quoted form for 0.71, and keep it there
>>>> for at least until SpamAssassin 3.4 really hits the streets;
>>>> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
>>>>
>>>> ...
>>>> Looking back at what I just wrote, the #2 looks to me
>>>> the most appealing.
>>> +1. Can you reach out to them and just ask for perhaps 3 months?
>>>
>>> Regards,
>>> KAM
>
>
> --
> *Kevin A. McGrail*
> President
>
> Peregrine Computer Consultants Corporation
> 3927 Old Lee Highway, Suite 102-C
> Fairfax, VA 22030-2422
>
> http://www.pccc.com/
>
> 703-359-9700 x50 / 800-823-8402 (Toll-Free)
> 703-359-8451 (fax)
> KMcGrail@PCCC.com <ma...@pccc.com>
>
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/24/2013 7:34 AM, Willem Toorop wrote:
> Hi Spamassassin developers,
>
> This message was from more than a year ago. We would like to revert our
> temporary workaround in Net::DNS. Is it safe now? Is a Spamassassin
> released which works without the workaround?
>
> Best regards,
>
> -- Willem
Willem,
Could you give us one more month please? You are right it has been far
too long and I'm working on this now.
Regards,
KAM
>
> op 14-12-12 18:37, Kevin A. McGrail schreef:
>> On 12/14/2012 12:33 PM, Mark Martinec wrote:
>>> 2) ask the Net::DNS folks if they'd be willing to put back the
>>> unconditionally double-quoted form for 0.71, and keep it there
>>> for at least until SpamAssassin 3.4 really hits the streets;
>>> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
>>>
>>> ...
>>> Looking back at what I just wrote, the #2 looks to me
>>> the most appealing.
>> +1. Can you reach out to them and just ask for perhaps 3 months?
>>
>> Regards,
>> KAM
--
*Kevin A. McGrail*
President
Peregrine Computer Consultants Corporation
3927 Old Lee Highway, Suite 102-C
Fairfax, VA 22030-2422
http://www.pccc.com/
703-359-9700 x50 / 800-823-8402 (Toll-Free)
703-359-8451 (fax)
KMcGrail@PCCC.com <ma...@pccc.com>
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Benny Pedersen <me...@junc.eu>.
Kevin A. McGrail skrev den 2014-01-11 16:52:
> Not sure what you mean. Did you not see the post with the link?
yep just to late after my own post
> http://people.apache.org/~kmcgrail/devel/Mail-SpamAssassin-3.4.0-rc4.tar.bz2
i use this link in my gentoo ebuilds, 3.4 is not yet in gentoo so i
create my own problem :)
> Also, if you are looking for the Changes, see
> http://svn.apache.org/repos/asf/spamassassin/trunk/Changes because
> it's a proposed announcement so it isn't yet tagged as a branch.
ok
> Will be creating rc5 shortly to address bug 6990 with Redis.
ok
Re: rc issues was Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Benny Pedersen <me...@junc.eu>.
Kevin A. McGrail skrev den 2014-01-12 22:46:
> But you asked about a locale pl. Any output you can share would help
> because I do not see where sa is setting a locale of pl
t/lang_pl_tests.t change env from pl to POSIX, it refer to locale -m
output ?
there is no pl here in my locale -m output so this seems to be the
problem, i have tryed with POSIX and it worked with that test
pl_PL is valid glibc locale, but pl is not
unpack tarball
perl Makefile.PL
make disttest
there i go :)
resolved:
#!/usr/bin/perl
use lib '.'; use lib 't';
use SATest; sa_t_init("lang_pl_tests");
use Test; BEGIN { plan tests => 1 };
#
---------------------------------------------------------------------------
%patterns = (
q{ X-Spam-Status: }, 'didnt_hang_at_least',
);
$ENV{'PERL_BADLANG'} = 0; # Sweep problems under the rug
$ENV{'LANGUAGE'} = 'pl_PL';
$ENV{'LC_ALL'} = 'POSIX';
sarun ("-L -t < data/nice/004", \&patterns_run_cb);
ok_all_patterns();
no ?
Re: rc issues was Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
But you asked about a locale pl. Any output you can share would help because I do not see where sa is setting a locale of pl
Regards,
KAM
Benny Pedersen <me...@junc.eu> wrote:
>Kevin A. McGrail skrev den 2014-01-12 16:36:
>
>> Need more context. The only setlocate in the SA source that I can
>> find is in spamd:
>
>yes its there
>
>> spamd/spamd.raw:POSIX::setlocale(LC_TIME,'C');
>> Where are you seeing this?
>
>fail is seen when doing perl test stagebuild, this is done in gentoo
>ebuild create with "ebuild spamassassin test" , if that fails i will
>remove the local ebuild, but it pass, with the above as warning
>
>this ebuild is unofficial so it might be my own fault :(
Re: rc issues was Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Benny Pedersen <me...@junc.eu>.
Kevin A. McGrail skrev den 2014-01-12 16:36:
> Need more context. The only setlocate in the SA source that I can
> find is in spamd:
yes its there
> spamd/spamd.raw:POSIX::setlocale(LC_TIME,'C');
> Where are you seeing this?
fail is seen when doing perl test stagebuild, this is done in gentoo
ebuild create with "ebuild spamassassin test" , if that fails i will
remove the local ebuild, but it pass, with the above as warning
this ebuild is unofficial so it might be my own fault :(
rc issues was Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 1/11/2014 12:28 PM, Benny Pedersen wrote:
>
> only one possible fail here (warning)
>
> could not setlocale LC_ALL (pl)
>
> is pl valid glibc locale ?
Need more context. The only setlocate in the SA source that I can find
is in spamd:
spamd/spamd.raw:POSIX::setlocale(LC_TIME,'C');
Where are you seeing this?
Regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 1/11/2014 8:06 AM, Benny Pedersen wrote:
> Kevin A. McGrail skrev den 2014-01-11 07:33:
>
>> Just a quick note to thank you for your patience and let you know we
>> put out a rc4 for release 3.4.0 today!
>
> talk is free, where is the link ? :=)
Not sure what you mean. Did you not see the post with the link?
http://mail-archives.apache.org/mod_mbox/spamassassin-users/201401.mbox/browser
For example:
http://people.apache.org/~kmcgrail/devel/Mail-SpamAssassin-3.4.0-rc4.tar.bz2
Also, if you are looking for the Changes, see
http://svn.apache.org/repos/asf/spamassassin/trunk/Changes because it's
a proposed announcement so it isn't yet tagged as a branch.
Will be creating rc5 shortly to address bug 6990 with Redis.
Regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Benny Pedersen <me...@junc.eu>.
Kevin A. McGrail skrev den 2014-01-11 07:33:
> Just a quick note to thank you for your patience and let you know we
> put out a rc4 for release 3.4.0 today!
talk is free, where is the link ? :=)
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
Hi Willem,
Just a quick note to thank you for your patience and let you know we put
out a rc4 for release 3.4.0 today!
regards,
KAM
On 12/24/2013 7:34 AM, Willem Toorop wrote:
> Hi Spamassassin developers,
>
> This message was from more than a year ago. We would like to revert our
> temporary workaround in Net::DNS. Is it safe now? Is a Spamassassin
> released which works without the workaround?
>
> Best regards,
>
> -- Willem
>
> op 14-12-12 18:37, Kevin A. McGrail schreef:
>> On 12/14/2012 12:33 PM, Mark Martinec wrote:
>>> 2) ask the Net::DNS folks if they'd be willing to put back the
>>> unconditionally double-quoted form for 0.71, and keep it there
>>> for at least until SpamAssassin 3.4 really hits the streets;
>>> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
>>>
>>> ...
>>> Looking back at what I just wrote, the #2 looks to me
>>> the most appealing.
>> +1. Can you reach out to them and just ask for perhaps 3 months?
>>
>> Regards,
>> KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Willem Toorop <wi...@nlnetlabs.nl>.
Hi Spamassassin developers,
This message was from more than a year ago. We would like to revert our
temporary workaround in Net::DNS. Is it safe now? Is a Spamassassin
released which works without the workaround?
Best regards,
-- Willem
op 14-12-12 18:37, Kevin A. McGrail schreef:
> On 12/14/2012 12:33 PM, Mark Martinec wrote:
>> 2) ask the Net::DNS folks if they'd be willing to put back the
>> unconditionally double-quoted form for 0.71, and keep it there
>> for at least until SpamAssassin 3.4 really hits the streets;
>> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
>>
>> ...
>> Looking back at what I just wrote, the #2 looks to me
>> the most appealing.
>
> +1. Can you reach out to them and just ask for perhaps 3 months?
>
> Regards,
> KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/14/2012 12:33 PM, Mark Martinec wrote:
> 2) ask the Net::DNS folks if they'd be willing to put back the
> unconditionally double-quoted form for 0.71, and keep it there
> for at least until SpamAssassin 3.4 really hits the streets;
> meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
>
> ...
> Looking back at what I just wrote, the #2 looks to me
> the most appealing.
+1. Can you reach out to them and just ask for perhaps 3 months?
Regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
> This means if we have to
> drop support for older Net::DNS, I would support that as well.
No need to, this is not in question.
> Right now, as I understand it:
>
> A) you've already patched sa-update in trunk to handle this and it works
> with older and newer Net::DNS versions
Right.
> B) Rather than change our backend automated DNS to add the kludgey
> entries and hope they work and aren't destroyed in the DNS ecosystem, if
> the community demand exists, we push a new 3.3 release with the same patch.
>
> Overall, 3.4 has been far too long in the making and contains a
> wonderful number of bug fixes. Running 3.3 at this point is just not
> what I would recommend to anyone.
I agree that pushing 3.4 out the door as soon as possible
is the right strategy. Trying to make a 3.3.3 release
seems to me an unnecessary effort.
Regarding the very shory-term future, I see the following choices
(in no particular order):
1) hope that Linux/Unix distros pick up and apply the small patch
to sa-update; contact them if necessary; meanwhile answer concerns
on the mailing list, pointing people to apply the patch to SA
or to downgrade their Net::DNS to 0.68;
2) ask the Net::DNS folks if they'd be willing to put back the
unconditionally double-quoted form for 0.71, and keep it there
for at least until SpamAssassin 3.4 really hits the streets;
meanwhile suggest people to switch Net::DNS to 0.71 or to 0.68;
3) make a 3.3.3 release with just this small patch applied;
4) go with the quick-fix / kludge workaround;
Looking back at what I just wrote, the #2 looks to me
the most appealing.
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Todd Rinaldo <to...@cpanel.net>.
On Dec 14, 2012, at 6:10 AM, Mark Martinec <Ma...@ijs.si> wrote:
>> I think we will focus on 3.4 released with the fix.
>> People using 3.3 can apply the patch, yes?
>
> This is certainly an option. Timeliness of this depends much
> on the responsiveness of popular Unix/Linux distributions,
> I'd expect that a majority of end-users are reluctant
> to apply local patches if they get their software prepackaged
> through some distro.
Yeah. I suspect anyone looking for a 3.3 patch is going to find it. It would be nice to see 3.4 for christmas, though.
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/14/2012 7:10 AM, Mark Martinec wrote:
>> I think we will focus on 3.4 released with the fix.
>> People using 3.3 can apply the patch, yes?
> This is certainly an option. Timeliness of this depends much
> on the responsiveness of popular Unix/Linux distributions,
> I'd expect that a majority of end-users are reluctant
> to apply local patches if they get their software prepackaged
> through some distro.
>
>> Comment 3 on bug 6872? The adding @ and space?
>> I think it's a horrible kludge, sorry.
> It looks horrible, but comes from the RFC 1035 section 5.1
> syntax specs for a zone file syntax:
>
> <character-string> is expressed in one or two ways: as a contiguous set
> of characters without interior spaces, or as a string beginning with a "
> and ending with a ". Inside a " delimited string any character can
> occur, except for a " itself, which must be quoted using \ (back slash).
>
> It is very unfortunate that the Net::DNS chose the syntax of a zone
> file for its API (which need not be constrained to 7-bit ascii text).
> It is also very unfortunate that SpamAssassin in many places
> ignores this fact and assumes an 8-bit clean API to Net::DNS.
>
> By a lucky coincidence, in most cases (plain text labels and data)
> both forms happen to be the same (apart from double quotes).
>
> The RFC 1035 section 5.1 text makes double quotes optional
> around a simple text with no spaces, and this is the reason
> the Net::DNS chose now to drop this optional double quotes
> unless the string contains funny characters.
>
> So the bug lies on our wrong assumption that Net::DNS API
> provides access to data with no zone file encodings.
>
> Adding a space or an @ makes Net::DNS choose the double-quote
> form, which is what sa-update 3.3 unfortunately expects.
>
> So yes, it is a kludge, but not a freevolous hack, but
> something that comes directly from RFC 1035.
Thanks. This is really well researched and defends Net::DNS's choice to
make this change quite well.
From SA's perspective, I believe we should follow the RFC and release
versions that work with the newer Net::DNS. This means if we have to
drop support for older Net::DNS, I would support that as well.
Right now, as I understand it:
A) you've already patched sa-update in trunk to handle this and it works
with older and newer Net::DNS versions
B) Rather than change our backend automated DNS to add the kludgey
entries and hope they work and aren't destroyed in the DNS ecosystem, if
the community demand exists, we push a new 3.3 release with the same patch.
Overall, 3.4 has been far too long in the making and contains a
wonderful number of bug fixes. Running 3.3 at this point is just not
what I would recommend to anyone.
Regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
> I think we will focus on 3.4 released with the fix.
> People using 3.3 can apply the patch, yes?
This is certainly an option. Timeliness of this depends much
on the responsiveness of popular Unix/Linux distributions,
I'd expect that a majority of end-users are reluctant
to apply local patches if they get their software prepackaged
through some distro.
> Comment 3 on bug 6872? The adding @ and space?
> I think it's a horrible kludge, sorry.
It looks horrible, but comes from the RFC 1035 section 5.1
syntax specs for a zone file syntax:
<character-string> is expressed in one or two ways: as a contiguous set
of characters without interior spaces, or as a string beginning with a "
and ending with a ". Inside a " delimited string any character can
occur, except for a " itself, which must be quoted using \ (back slash).
It is very unfortunate that the Net::DNS chose the syntax of a zone
file for its API (which need not be constrained to 7-bit ascii text).
It is also very unfortunate that SpamAssassin in many places
ignores this fact and assumes an 8-bit clean API to Net::DNS.
By a lucky coincidence, in most cases (plain text labels and data)
both forms happen to be the same (apart from double quotes).
The RFC 1035 section 5.1 text makes double quotes optional
around a simple text with no spaces, and this is the reason
the Net::DNS chose now to drop this optional double quotes
unless the string contains funny characters.
So the bug lies on our wrong assumption that Net::DNS API
provides access to data with no zone file encodings.
Adding a space or an @ makes Net::DNS choose the double-quote
form, which is what sa-update 3.3 unfortunately expects.
So yes, it is a kludge, but not a freevolous hack, but
something that comes directly from RFC 1035.
Mark
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by "Kevin A. McGrail" <KM...@PCCC.com>.
On 12/13/2012 8:22 PM, Mark Martinec wrote:
> Any thoughts on the workaround as proposed by Willem?
>
> Anybody familiar enough with the DNS infrastructure
> of the project to dare implementing it?
>
> Mark
Comment 3 on bug 6872? The adding @ and space? I think it's a horrible
kludge, sorry.
I think we will focus on 3.4 released with the fix. People using 3.3
can apply the patch, yes?
Regards,
KAM
Re: [Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by Mark Martinec <Ma...@ijs.si>.
Any thoughts on the workaround as proposed by Willem?
Anybody familiar enough with the DNS infrastructure
of the project to dare implementing it?
Mark
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #10 from Mark Martinec <Ma...@ijs.si> ---
Created attachment 5118
--> https://issues.apache.org/SpamAssassin/attachment.cgi?id=5118&action=edit
A minimalistic patch to sa-update for SA 3.3
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #12 from Mark Martinec <Ma...@ijs.si> ---
Meanwhile the Net::DNS 0.71 has been released, temporarily reverting the
change that broke sa-update, until we get our 3.4.0 out - thank you Willem
and the rest of the NLnet Lab crew!
This is from Net::DNS 0.71 release notes:
**** 0.71 Dec 15, 2012
Temporary workaround rt.cpan.org #81760
The rdatastr method for TXT RRs will return unconditionally
quoted rdata fields to work around an issue with updating
SpamAssassin rules. This workaround will be reverted after
release of a version of SpamAssassin which resolves the issue.
After some offline discussion with Willem (me complaining about
the difficulties when a caller wants to access data strings from
TXT records with multiple RDATA <character-string> fields, while
also avoiding the Net::DNS habit of using a zone-file format
encodings in the API, only to have the caller go into trouble
of undoing the unnecessary encoding) - Willem suggested the
following:
> > So what can I do? Check a Net::DNS version and use char_str_list()
> > with older Net::DNS, and txtdata() with newer???
>
> I promise to discuss this thoroughly with Dick and Olaf and report back
> with a statement on which you can build and depend.
>
> In the mean time, I encourage you to use txtdata for getting the values
> of <version>.updates.spamassassin.org and
> mirros.updates.spamassassin.org. As those records have only a single
> rdata field, txtdata would return the same value since Net::DNS 0.34.
This sounds like a good advise, especially since Net::DNS 0.34 already
happen to be the lowest supported version of Net::DNS in SpamAssassin.
So I'm folding in a couple of changes to implement it. Will attach
the diff in the next posting. The following new comments in the code
and a few samples of the change should make the reasoning clearer,
I hope:
- $log = $answer->rdatastr;
- $log =~ s/^"(.*)"\z/$1/s;
+ # txtdata returns a non- zone-file-format encoded result, unlike rdatastr;
+ # avoid space-separated RDATA <character-string> fields if possible,
+ # txtdata provides a list of strings in a list context since Net::DNS 0.69
+ $log = join('',$answer->txtdata);
+ # NOTE: $rr->rdatastr returns the result encoded in a DNS zone file
+ # format, i.e. enclosed in double quotes if a result contains whitespace
+ # (or other funny characters), and may use \DDD encoding or \X quoting as
+ # per RFC 1035. Using $rr->txtdata instead avoids this unnecessary encoding
+ # step and a need for decoding by a caller, returning an unmodified string.
+ # Caveat: in case of multiple RDATA <character-string> fields contained
+ # in a resource record (TXT, SPF, HINFO), starting with Net::DNS 0.69
+ # the $rr->txtdata in a list context returns these strings as a list.
+ # The $rr->txtdata in a scalar context always returns a single string
+ # with <character-string> fields joined by a single space character as
+ # a separator. The $rr->txtdata in Net::DNS 0.68 and older returned
+ # such joined space-separated string even in a list context.
+ #
+ # From Net::DNS maintainers (Willem Toorop, NLnet Labs):
+ # I encourage you to use txtdata for getting the values of
+ # <version>.updates.spamassassin.org and mirros.updates.spamassassin.org.
+ # As those records have only a single rdata field, txtdata would return
+ # the same value since Net::DNS 0.34.
- my $text = $rr->rdatastr;
- if (defined $text && $text ne '') {
- local($1);
- $text =~ s/^"(.*)"\z/$1/s;
- push(@result,$text);
- }
+ # scalar context!
+ my $text = $rr->UNIVERSAL::can('txtdata') ? $rr->txtdata :
$rr->rdatastr;
+ push(@result,$text) if defined $text && $text ne '';
Trunk:
Bug 6872 - replace $rr->rdatastr with $rr->txtdata where appropriate
Sending lib/Mail/SpamAssassin/Dns.pm
Sending lib/Mail/SpamAssassin/DnsResolver.pm
Sending lib/Mail/SpamAssassin/Plugin/AskDNS.pm
Sending lib/Mail/SpamAssassin/Plugin/URIDNSBL.pm
Sending sa-update.raw
Committed revision 1425068.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #2 from Mark Martinec <Ma...@ijs.si> ---
Created attachment 5115
--> https://issues.apache.org/SpamAssassin/attachment.cgi?id=5115&action=edit
a fix to sa-update, and some other DNS-related tidbits
trunk:
r1418317 | mmartinec | 2012-12-07 15:04:22 +0100 (Fri, 07 Dec 2012)
- Net::DNS 0.69 $rr->rdatastr no longer quotes returned TXT
([rt.cpan.org #81760])
- some other DNS-related cosmetics
- more RR types in AskDNS.pm
Copied from a mailing list:
Mark Martinec wrote:
> Net::DNS has broken sa-update -
> https://rt.cpan.org/Ticket/Display.html?id=81760
Interesting. The Net::DNS 0.68 docs claim:
man Net::DNS::RR::TXT
char_str_list
print "Individual <character-string> list: \n\t",
join("\n\t", $rr->char_str_list());
Returns a list of the individual <character-string> elements, as
unquoted strings. Used by TXT->rdatastr and TXT->rr_rdata.
NB: rdatastr will return quoted strings.
But the inserted quotation marks seem silly indeed.
> This patch to sa-update would work around Net::DNS 0.70. $1 should never be
> used anyways unless you're sure the regex matched. It's just asking for
> trouble.
Thanks, fixed now in trunk, r1418317.
Used the same approach/code for sa-update, as was properly used
in SpamAssassin/Dns.pm.
> Also local $1 makes me nervous. It could potentially behave
> different on different versions of perl depending on if bugs existed with
> local or $1.
It's the other way around: missing local($1) could behave differently
depending on a version of perl. Having global $n variables localized
before use makes them undefined and taint-free, and avoids potentially
polluting the same variable used by a caller when invoking some subroutine
or hitting an exception handler. This applies to any version of perl,
it makes a good practice.
Kevin A. McGrail wrote:
Thanks Mark for reviewing this. And Todd, good catch.
I agree with Mark's patch re: local but otherwise a good fix.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #4 from Todd Rinaldo <to...@cpanel.net> ---
If Net::DNS will not be putting the quotes back, It seems like this patch would
do well to go into a 3.3.3 branch with only this fix to be released to CPAN. Is
this in the works? Potentially all 3.3.2 users who have updated Net::DNS are
now broken. Also any new installs are potentially broken.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
Kevin A. McGrail <km...@pccc.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
CC| |kmcgrail@pccc.com
Resolution|--- |FIXED
--- Comment #18 from Kevin A. McGrail <km...@pccc.com> ---
I am closing this ticket because we are not going to backport to 3.3 and trunk
can handle the new versions which should give us enough time to handle the
change with 3.4.0's release.
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #9 from Willem Toorop <wi...@nlnetlabs.nl> ---
(In reply to comment #8 by Mark Martinec)
> I think adopting Willem's suggestion is a good (yet hackish) short term
> solution until the 3.4 can be released to a general public (noting that
> several sites are already using 3.4(trunk) in production).
Thanks! I hope you will apply it. Let me know if you will so I can release 0.71
with the other issues resolved (DKIM etc).
For smoothest transition I think it would be best to first update
mirrors.updates.spamassassin.org, and then wait for one hour (the TTL of the
records) before updating <version>.updates.spamassassin.org. That way you are
sure that if an sa-update gets a new version number from DNS, it can also get
the url for fetching MIRRORED.BY.
If you alter them at the same time you have a small risk that a resolver still
has the old value for mirrors.updates.spamassassin.org in its cache whilst the
<version>.updates.spamassassin.org is updated causing sa-update to fail with:
error: no mirror data available for channel updates.spamassassin.org
channel: MIRRORED.BY file location was not in DNS, channel failed
Best regards, Willem Toorop
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #17 from tycho.eggen@gmail.com ---
Apologies, my support ticket should have been filed as a PEBKAC issue.
*enables a certain update from crontab which was commented out for safety
reasons*
I thank you humbly for your time and effort.
I bid you a good night!
--
You are receiving this mail because:
You are the assignee for the bug.
[Bug 6872] Net::DNS 0.69 breaks sa-update
Posted by bu...@bugzilla.spamassassin.org.
https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6872
--- Comment #6 from Mark Martinec <Ma...@ijs.si> ---
Before returning to the original problem report, first a quick explanation
on what my first patch broke. The symptoms were noticed by very long nightly
runs of masschecks caused by unnecessarily long scan times. If there was
at least one unsuccessful DNS reply, it was ignored and the code then
waited until RBL timeout (default 15 seconds, often set to 5 seconds).
The culprit is the following section in the old code in DnsResolver.pm
(before any changes):
if (defined $packet &&
defined $packet->header &&
defined $packet->question &&
defined $packet->answer)
{ ... deal with a received DNS packet ... }
The problem here is that $packet->question and $packet->answer
return a list (possibly empty). The defined() of such list always
returns true, so the above test just wasted unnecessary cycles
to test the question and answer sections, but then ignored the
result.
My first patch tried to 'fix' this by doing what seemed to be
the intention, i.e. ignore a packet with an empty answer section.
But that is wrong, such a response still needs to hit its
callback routine and be removed from the list of waited-upon
queries. As it happens, the original (broken) code managed
to be doing the right thing, while the 'fixed' code broke it.
Now, the today's change hopefully does *the*right*thing*,
while also providing a more informative debugging.
--
You are receiving this mail because:
You are the assignee for the bug.