In testing in my lab, I've found significant issues using SpamAssassin
3.4.1 with Net::DNS 1.02 or later. Previously, I was using 0.81.
With Net::DNS 1.02 or 1.04, there is an 15 second+ delay in delivering
email. With debugging enabled for SA, we see the first delay here:
Dec 9 15:19:56 zre-ldap002 amavis[14134]: (14134-03) p.path
testus...@zre-ldap002.eng.zimbra.com: "P=p003,L=1,M=multipart/alternative |
P=p002,L=1/2,M=text/html,T=asc"
Dec 9 15:20:02 zre-ldap002 amavis[14134]: (14134-03) SA dbg: dns: select
timed out 1.000 s
Dec 9 15:20:02 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async: select
found no responses ready (t.o.=1.0)
Dec 9 15:20:02 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async:
queries completed: 0, started: 0
This 6 second delay is consistent. It is followed by a ton of SA dbg: dns
messages.
Eventually we get to:
Dec 9 15:20:11 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async:
escaping: lost or timed out requests or responses
Dec 9 15:20:11 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async:
aborting after 15.010 s, past original deadline: TXT,
askdns:TXT:_dmarc.zre-ldap002.eng.zimbra.com
Dec 9 15:20:11 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async:
aborting after 15.003 s, past original deadline: NO_DNS_FOR_FROM, DNSBL-A,
dns:A:zre-ldap002.eng.zimbra.com
Dec 9 15:20:11 zre-ldap002 amavis[14134]: (14134-03) SA dbg: async:
aborting after 15.002 s, past original deadline: NO_DNS_FOR_FROM, DNSBL-MX,
dns:MX:zre-ldap002.eng.zimbra.com
So this makes sense for 15 seconds, since that seems to be the timer
default. I've even patched SA with the two commits from Mark to workaround
the issues in 1.03 that were introduced. However, that doesn't change the
problem I see here, so there are still apparently issues with using current
versions of Net::DNS with SpamAssassin.
With Net::DNS 0.81, I never see any dns debug lines from SA:
Dec 9 14:48:45 zre-ldap002 amavis[428]: (00428-03) p.path
testus...@zre-ldap002.eng.zimbra.com: "P=p003,L=1,M=multipart/alternative |
P=p002,L=1/2,M=text/html,T=asc"
Dec 9 14:48:45 zre-ldap002 postfix/amavisd/smtpd[3048]: 7F1021160815:
client=localhost[127.0.0.1]
Dec 9 14:48:45 zre-ldap002 postfix/cleanup[2268]: 7F1021160815:
message-id=<1744051725.1.1449694124920.javamail.zim...@zre-ldap002.eng.zimbra.com>
Dec 9 14:48:45 zre-ldap002 postfix/amavisd/smtpd[3048]: disconnect from
localhost[127.0.0.1] ehlo=1 mail=11 rcpt=11 data=11 noop=1 quit=1
commands=36
Dec 9 14:48:45 zre-ldap002 postfix/qmgr[2267]: 7F1021160815:
from=<testus...@zre-ldap002.eng.zimbra.com>, size=2457, nrcpt=1 (queue
active)
Dec 9 14:48:45 zre-ldap002 postfix/smtp[3010]: 35C101160816:
to=<testus...@zre-ldap002.eng.zimbra.com>,
relay=127.0.0.1[127.0.0.1]:10032, delay=0.34, delays=0.07/0/0/0.27,
dsn=2.0.0, status=sent (250 2.0.0 from MTA(smtp:[127.0.0.1]:10025): 250
2.0.0 Ok: queued as 7F1021160815)
Dec 9 14:48:45 zre-ldap002 postfix/qmgr[2267]: 35C101160816: removed
Dec 9 14:48:45 zre-ldap002 postfix/lmtp[3087]: 7F1021160815:
to=<testus...@zre-ldap002.eng.zimbra.com>,
relay=zre-ldap002.eng.zimbra.com[10.137.242.52]:7025, delay=0.25,
delays=0.03/0/0.1/0.12, dsn=2.1.5, status=sent (250 2.1.5 Delivery OK)
Instead, I get immediate delivery.
Generally, I'd recommend against using current versions of Net::DNS until
this can get sorted out.
--Quanah
--
Quanah Gibson-Mount
Platform Architect
Zimbra, Inc.
--------------------
Zimbra :: the leader in open source messaging and collaboration
