Bug 56968
| Summary: | Disabling improper SMTP '4xx' processing by mailman, and workaround | ||
|---|---|---|---|
| Product: | [Retired] Red Hat Raw Hide | Reporter: | R P Herrold <herrold> |
| Component: | mailman | Assignee: | John Dennis <jdennis> |
| Status: | CLOSED WONTFIX | QA Contact: | |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 1.0 | ||
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2003-03-13 21:51:22 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
Hmm... I'm not sure what to do with this bug, its pretty old, mailman has moved on, current version is 2.1.1 and I'm not sure I even have the ability to reproduce it. I'm going to close it for now unless somehow howls. |
Extended Summary: ================================ mailman delivery outbound queue can get 'stuck' when the next MTA returns a 4xx series transient failure. mailman should PROPERLY go on the the next piece in its run queue, but instead improperly stops processing the queue at that point. -------------------------------------------------- Latest RawHide version: #[herrold@stones mail]$ rpm -q mailman mailman-2.0.7-1 -------------------------------------------------- The Issue =================================== 1. This was noticed, when caused by remote site DNS issues: [herrold@stones sendmail]$ dig cimetric.com mx ; <<>> DiG 9.1.0 <<>> cimetric.com mx ;; global options: printcmd ;; connection timed out; no servers could be reached ... that is, no MX is being returned, and indeed the remote authoratative nameserver, for whatever is not responding 2. The domain appears to not be expired ... [herrold@stones sendmail]$ truewho cimetric.com Test of domain: cimetric.com Date: Sat Dec 1 11:37:58 EST 2001 ================================================== whois cimetric.com.com [whois.register.com] Organization: Cimetric Commerce, Inc. <snip> Domain Name: CIMETRIC.COM Created on..............: Fri, May 19, 2000 Expires on..............: Sun, May 19, 2002 Record last updated on..: Tue, Jul 31, 2001 Domain servers in listed order: NS1.NCOMGROUP.COM 38.196.141.10 NS2.NCOMGROUP.COM 38.196.141.20 3. Cross check to show that WE do not havee LOCAL resolver issues: [herrold@stones sendmail]$ dig redhat.com mx ; <<>> DiG 9.1.0 <<>> redhat.com mx ;; global options: printcmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 20953 ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 3, ADDITIONAL: 6 ;; QUESTION SECTION: ;redhat.com. IN MX ;; ANSWER SECTION: redhat.com. 86400 IN MX 10 mail.redhat.com. -------------- Detection ========================= 4. How found: herrold@stones logs]$ pwd ; tail smtp-failure /var/mailman/logs Dec 01 11:32:25 2001 (4864) 450 paul.timmerman (4.7.1 <paul.timmerman>... Can not check MX records for recipient host cimetric.com) Dec 01 11:33:40 2001 (4864) 450 paul.timmerman (4.7.1 <paul.timmerman>... Can not check MX records for recipient host cimetric.com) ... and of course no list mail was flowing ... TWO PORT FIX ================================== 5. As a short term workaround, I added the hosed domain to /etc/mail/local [herrold@stones mail]$ pwd ; tail local-host-names /etc/mail # # # 011201 penguin.lvcablemodem.com cimetric.com # 6. -AND- loosened my /etc/mail/access to accept ALL content proffered to sendmail from the local mailman (this should probably be added by default anyway, if not present): # ####################################################### # # RPH 011201 # # Some sendmail 8.10.x and later mail anti-spam rules can cause # mailman proffered content to be temporarily declined, and # a smtp '450' status is handed back to mailman -- # # At this point mailman, through ver 2.0.7 gets wedged and # blocked from passing along _other_ content to differing # addressees ... It is really a mailman issue as to how # it is handling '4xx' messages, but ... # # This entry into /etc/mail/accesss should unquestionably # allow mailman to hand mail off to the local sendmail # instance, and let the MTA worry about further delivery # mailman@localhost RELAY # ####################################################