Emails are stuck in deferred queue

[Brieuc]

Hi,

Our Zimbra platform worked well until a few days ago when all the message stop being delivered to the mailboxes to be added to the deferred queue.

Here is the error message I got :

status=deferred (conversation with mail.mydomain.com[xxx.xxx.xxx.xxx] timed out while sending end of data -- message may be sent more than once)

I read different post in the forum where issues seems to be the same as mine, but I was not able to figure out my issue with the answer to those post.

I checked with the "Incoming Mail Problems" wiki but it didn't solved the issue.

Here are some command result for troubleshooting :

host `hostname`

mydomain.com has address 192.168.xxx.xxx
mydomain.com mail is handled by 10 mydomain.com.

/etc/hosts

127.0.0.1   localhost.localdomain   localhost
192.168.xxx.xxx  mydomain.com clamtc02
#192.168.xxx.xxx clamtc02.mydomain.com clamtc02

172.26.252.xxx   nlasav02    nlasav02.mydomain.com
172.26.252.yyy   nlasav04    nlasav04.mydomain.com
172.26.252.zzz   nlasav05    nlasav05.mydomain.com
172.16.xxx.xxx   nvasav01    nvasav01.mydomain.com
172.16.yyy.yyy   nvasav03     nvasav03.mydomain.com

cat /etc/resolv.conf

search  internal_domain.com
nameserver 172.26.xxx.xxx
nameserver 172.26.xxx.xxx

dig mydomain.com MX

; <<>> DiG 9.3.4-P1 <<>> mydomain.com MX
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 56947
;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;mydomain.com.         IN      MX

;; ANSWER SECTION:
mydomain.com.  6718    IN      MX      10 smtp2.mydomain.com.
mydomain.com.  6718    IN      MX      10 smtp.mydomain.com.

;; Query time: 338 msec
;; SERVER: 172.26.xxx.xxx#53(172.26.xxx.xxx)
;; WHEN: Mon Jul 16 18:56:39 2012
;; MSG SIZE  rcvd: 82

dig mydomain.com any

; <<>> DiG 9.3.4-P1 <<>> mydomain.com any
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 36709
;; flags: qr rd ra; QUERY: 1, ANSWER: 8, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;mydomain.com.         IN      ANY

;; ANSWER SECTION:
mydomain.com.  21600   IN      A       212.53.xxx.xxx
mydomain.com.  21600   IN      SOA     ns1.netnames.net. hostmaster.netnames.net. 2012012416 166066 83033 1162462 515586
mydomain.com.  21600   IN      NS      ns2.netnames.net.
mydomain.com.  21600   IN      NS      ns6.netnames.net.
mydomain.com.  21600   IN      NS      ns5.netnames.net.
mydomain.com.  21600   IN      NS      ns1.netnames.net.
mydomain.com.  21600   IN      MX      10 smtp.mydomain.com.
mydomain.com.  21600   IN      MX      10 smtp2.mydomain.com.

;; Query time: 58 msec
;; SERVER: 172.26.xxx.xxx#53(172.26.xxx.xxx)
;; WHEN: Mon Jul 16 19:01:01 2012
;; MSG SIZE  rcvd: 229

Here are other settings changes we tried :

• Increase / Decrease the MTU
• I increased the lmtp_destination_concurrency_limit (from 20 to 40)

One last information :
I am not able to telnet mydomain.com 10025
I am able to telnet localhost 10025

So here I am, with no other solution to investigate.

Does anyone has a clue for me ???

[phoenix]

The first thing to mention is that your hosts file is totally incorrect, go to the 'Verify....' section of the Split DNS wiki article (or search the forums) for the correct format and change the file. You should also check if your certificates have expired and recreate them if they have. You should always give details of the ZCS version you're running by updating your forum profile with the output of the following command:

zmcontrol -v

For the moment you should post that information in this thread and you should also update your forum profile with that information.

[schtipoun]

Hi Phoenix,

Thanks for your quick reply!

I'm working with Brieuc on this. Here is the version we are using: Release 7.1.4_GA_2555.RHEL5_64_20120105094627 RHEL5_64 FOSS edition.

I'll go through the Split DNS Wiki to modify the file.

Do you know how I can check if the certificate is expired?

Thanks,
Antoine

[phoenix]

Do you know how I can check if the certificate is expired?

Administration Console and CLI Certificate Tools - Zimbra :: Wiki

[Brieuc]

Hi Bill,

Thanks for the advise, I just updated my profile with the ZCS version.
As schtipoun said, we are using Release 7.1.4_GA_2555.RHEL5_64_20120105094627 RHEL5_64 FOSS edition.

Regarding the certificate, we regenerated the SSL certificates and on the UI certificates are ok.

We corrected the hosts file (our specificity is that we changed the hostname by tech.mail.domain.com). As a reminder it worked for almost a year, I mean that hostname seems to work like that.

127.0.0.1       localhost.localdomain   localhost
192.168.104.25  tech.mail.domain.com clamtc02
192.168.104.25  clamtc02.internal_domain_name.net

172.26.252.85   nlasav02    nlasav02.internal_domain_name.net
172.26.252.88   nlasav04    nlasav04.internal_domain_name.net
172.26.252.84   nlasav05    nlasav05.internal_domain_name.net
172.16.2.134   nvasav01    nvasav01.internal_domain_name.net
172.16.3.32   nvasav03     nvasav03.internal_domain_name.net

Brieuc

[Brieuc]

Hi,

Our issue is now solved, here are the details (if anyone encountered the same issue) :

One of the mailbox dedicated for application batch received around 100 000 mails per day.
The user set up a forwarding filters, on all messages received within the mailbox, to 12 email addresses (not hosted on Zimbra). Some of those address did not exist.
We supposed that for each message received and forward, the mailbox received 3 or 4 "invalid recipient" message, then once again it forward those messages and create a kind of loop.

The "loop" caused the server to manage to many messages and LTMP just stopped accepting message from Postfix.

Brieuc

P.S.: how to report this thread as solved ?