Re: Exim4 behaviour when long term failure of outgoing address
On 10/10/12 17:00, Alan Chandler wrote: On 10/10/12 09:49, Chris Davies wrote: 4. Failure (bounce) message to root@avalon is being lost - and this is the issue at stake Absolutely correct. I think I have discovered - at least part of the problem - maybe the whole thing. The inaddr.arpa address for the IP address of my virtual server resolved to avalon.hartley-web.com a domain I dropped somewhile ago. It should be updating itself to avalon.hartley-consultants.com as we speak. Also - I think I confused both Lee and yourself with the use of 77.96.120.60 as a relay domain I don't actually pass through this virtual server normally - I normally use my ISPs normal mail server. This was just so it was already pre-configured IF I SHOULD NEED IT. Thanks for everyone's help on this issue. I decided, on my virtual machine, to set up Postfix to become the MX for the domains I was supporting and to use the virtual alias capability to forward mail on to real mail addresses. It is working great - so I don't need to worry over this issue anymore. -- Alan Chandler http://www.chandlerfamily.org.uk -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/5077cad8.1080...@chandlerfamily.org.uk
Re: Exim4 behaviour when long term failure of outgoing address
On 07/10/12 22:31, Chris Davies wrote: Alan Chandlera...@chandlerfamily.org.uk wrote: I am using Debian Squeeze on a virtual machine that I lease. It has exim4 (light) version as its mail server. - its name is avalon.hartley-consultants.com However, it looks to me like its trying to send a failure e-mail to me locally somehow. 2012-10-05 07:42:09 1TK1bf-Mx-0C= r...@avalon.hartley-consultants.com U=root P=local S=389 2012-10-05 07:42:09 1TK1bf-Mx-0C ** i...@mynewdomain.com R=dnslookup T=remote_smtp: retry time not reached for any host after a long failure period 2012-10-05 07:42:09 1TK1bt-N0-DT remote host address is the local host: avalon.hartley-consultants.com It's difficult to tell without knowing the precise setup on the machine, but this looks like you've aliased root to the offsite address i...@mynewdomain.com, but then you've got an entry somewhere that tells avalon that it *is* mynewdomain.com. This could be an entry in /etc/hosts, an MX or A record in DNS, or some fancy aliasing somewhere associated with exim itself. Avalon accordingly tries to deliver to info, locally, and finds that this does not exist. Because it's already trying to deliver a bounce message it simply discards the bouncing bounce and aborts. Unfortunately, without knowing what mynewdomain.com really is, I can't run any non-local diagnostics for you. Unless mynewdomain.com really is yours, in which case you've got a configuration problem there because it's not accepting mail. Chris I'll try and be more specific The domain in question is virginiaparkinson.com and I am having particular difficulty with the domain name hosting company to get e-mail forwarding working with them. The virtual machine is a standard squeeze setup with my update-exim4.conf.conf dc_eximconfig_configtype='internet' dc_other_hostnames='' dc_local_interfaces='' dc_readhost='' dc_relay_domains='' dc_minimaldns='false' dc_relay_nets='127.0.0.1;77.96.120.60' dc_smarthost='' CFILEMODE='644' dc_use_split_config='true' dc_hide_mailname='' dc_mailname_in_oh='true' dc_localdelivery='mail_spool' (77.96.120.60 is my home ip address where my main mail server sits - because this is effectively a dynamic ip address I have to route all outgoing mail through a remote smtp server. Normally I use my ISPs mail server, but occassionally it becomes slow, or is blacklisted - and this allows me to rapidly switch to this machine to route outgoing mail through) /etc/aliases has root: alan.chand...@hartley-consultants.com in it the virtual machines ip address is 80.68.94.252 and both hartley-consultants.com and virginiaparkinson.com have this domain referencing 80.68.94.252 BUT their MX records both point else where. In fact hartley-consultants MX record points to 77.96.120.60, whereas virginiaparkinson.com mx records point somewhere completely different (at first trial at what seems a non existant mail server that was refusing connections) I am trying to fix that now. -- Alan Chandler http://www.chandlerfamily.org.uk -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/507518e9.9070...@chandlerfamily.org.uk
Re: Exim4 behaviour when long term failure of outgoing address
Alan Chandler a...@chandlerfamily.org.uk wrote: The domain in question is virginiaparkinson.com [..] The one you're trying to deliver to (i.e. mynewdomain.com in the original posting)? The virtual machine is a standard squeeze setup with my update-exim4.conf.conf dc_other_hostnames='' (77.96.120.60 is my home ip address where my main mail server sits - because this is effectively a dynamic ip address I have to route all outgoing mail through a remote smtp server. You have a mid-term problem here you're going to need to address: that if your 77 address is dynamic, each time it changes you'll have to update your VM's dc_relay_nets configuration entry. However, there are better solutions for this so I'll park it for now. (Use authentication from your home mail server to your VM.) /etc/aliases has root: alan.chand...@hartley-consultants.com That's on the VM? and both hartley-consultants.com and virginiaparkinson.com have this domain referencing 80.68.94.252 BUT their MX records both point else where. MX defines the delivery target, so that's what's relevant here. I'll try to reiterate the configuration and let's go from there (offlist if you like, since this isn't really a Debian issue). 1. Home server (name unknown, probably irrelevant) forwards email to your VM for onward delivery 2. VM is called avalon.hartley-consultants.com, configured as an Internet SMTP system using the standard Debian configuration 3. VM won't deliver to virginiaparkinson.com, but that issue is out of scope right now 4. Failure (bounce) message to root@avalon is being lost - and this is the issue at stake I think I'll stand by my original diagnosis, taht the key line is the exim4 error message, remote host address is the local host: avalon.hartley-consultants.com. Typically this means that exim4 doesn't know all its possible names. Specifically in your instance, exim4 is trying to deliver to avalon.hartley-consultants.com, but this resolves (/etc/hosts, MX, or A) to the local system. The problem is that exim4's local configuration doesn't include this name as one of the possible alternative local names (the Other destinations for which mail is accepted question). Add this hostname using dpkg-reconfigure exim4-config and see if that solves the immediate problem. Chris -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/6epfk9xrhl@news.roaima.co.uk
Re: Exim4 behaviour when long term failure of outgoing address
Alan Chandler a...@chandlerfamily.org.uk writes: I'll try and be more specific The domain in question is virginiaparkinson.com and I am having , [ dig -t mx virginiaparkinson.com ] | | ; DiG 9.8.1-P1 -t mx virginiaparkinson.com | ;; global options: +cmd | ;; Got answer: | ;; -HEADER- opcode: QUERY, status: NOERROR, id: 54135 | ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 2, ADDITIONAL: 2 | | ;; QUESTION SECTION: | ;virginiaparkinson.com. IN MX | | ;; ANSWER SECTION: | virginiaparkinson.com.86311 IN MX 10 mail2.virginiaparkinson.com. | virginiaparkinson.com.86311 IN MX 0 mail.virginiaparkinson.com. | | ;; AUTHORITY SECTION: | virginiaparkinson.com.86284 IN NS ns4.ukdnsservers.co.uk. | virginiaparkinson.com.86284 IN NS ns3.ukdnsservers.co.uk. | | ;; ADDITIONAL SECTION: | mail.virginiaparkinson.com. 86311 IN A 72.1.201.138 | mail2.virginiaparkinson.com. 86311 IN A 72.1.201.138 | | ;; Query time: 0 msec | ;; SERVER: 127.0.0.1#53(127.0.0.1) | ;; WHEN: Wed Oct 10 16:36:53 2012 | ;; MSG SIZE rcvd: 168 ` particular difficulty with the domain name hosting company to get e-mail forwarding working with them. The virtual machine is a standard squeeze setup with my update-exim4.conf.conf As suggested before, copy over the example configuration and adjust to your needs. The splitfile stuff doesn't give you a readable exim configuration. (77.96.120.60 is my home ip address where my main mail server sits - because this is effectively a dynamic ip address I have to route all outgoing mail through a remote smtp server. Normally I use my ISPs mail server, but occassionally it becomes slow, or is blacklisted - and this allows me to rapidly switch to this machine to route outgoing mail through) So how do you get incoming mail? The MX records tell MTAs to hand it over to 72.1.201.138 rather than 77.96.120.60. /etc/aliases has root: alan.chand...@hartley-consultants.com , [ dig -t mx hartley-consultants.com ] | | ; DiG 9.8.1-P1 -t mx hartley-consultants.com | ;; global options: +cmd | ;; Got answer: | ;; -HEADER- opcode: QUERY, status: NOERROR, id: 18241 | ;; flags: qr rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 2, ADDITIONAL: 0 | | ;; QUESTION SECTION: | ;hartley-consultants.com. IN MX | | ;; ANSWER SECTION: | hartley-consultants.com. 86400IN MX 0 chandlerfamily.org.uk. | hartley-consultants.com. 86400IN MX 10 chandlerfamily.org.uk. | | ;; AUTHORITY SECTION: | hartley-consultants.com. 86400IN NS ns4.ukdnsservers.co.uk. | hartley-consultants.com. 86400IN NS ns3.ukdnsservers.co.uk. | | ;; Query time: 221 msec | ;; SERVER: 127.0.0.1#53(127.0.0.1) | ;; WHEN: Wed Oct 10 16:42:37 2012 | ;; MSG SIZE rcvd: 146 ` , [ dig -t a chandlerfamily.org.uk | | ; DiG 9.8.1-P1 -t a chandlerfamily.org.uk | ;; global options: +cmd | ;; Got answer: | ;; -HEADER- opcode: QUERY, status: NOERROR, id: 30086 | ;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 2, ADDITIONAL: 2 | | ;; QUESTION SECTION: | ;chandlerfamily.org.uk. IN A | | ;; ANSWER SECTION: | chandlerfamily.org.uk.86358 IN A 77.96.120.60 | | ;; AUTHORITY SECTION: | chandlerfamily.org.uk.86358 IN NS ns1.ukdnsservers.co.uk. | chandlerfamily.org.uk.86358 IN NS ns2.ukdnsservers.co.uk. | | ;; ADDITIONAL SECTION: | ns1.ukdnsservers.co.uk. 86358 IN A 72.1.201.150 | ns2.ukdnsservers.co.uk. 86358 IN A 72.1.216.98 | | ;; Query time: 0 msec | ;; SERVER: 127.0.0.1#53(127.0.0.1) | ;; WHEN: Wed Oct 10 16:45:23 2012 | ;; MSG SIZE rcvd: 139 ` the virtual machines ip address is 80.68.94.252 and both hartley-consultants.com and virginiaparkinson.com have this domain referencing 80.68.94.252 BUT their MX records both point else where. The above dig results is what my MTA would get. I think I already suggested that you probably have some DNS issues involved? In fact hartley-consultants MX record points to 77.96.120.60, whereas virginiaparkinson.com mx records point somewhere completely different To where are they supposed to point to? You'd be best off getting a static IP address and the corresponding MX entries so you can receive incoming mail. If you manage to get MX entries for a dynamic IP, you can get away with using a smarthost for the outgoing mail. That has the disadvantage that when your IP changes, MTAs may try to hand over mail to the old IP address until they become aware that the MX entry has changed. Whoever receives that IP address will be able to receive your mail then. You can use fetchmail instead. Unfortunately, that has some significant disadvantages. (at first trial at what seems a non existant mail server that was refusing connections) I am trying to fix that now. ,
Re: Exim4 behaviour when long term failure of outgoing address
On 10/10/12 09:49, Chris Davies wrote: 4. Failure (bounce) message to root@avalon is being lost - and this is the issue at stake Absolutely correct. I think I have discovered - at least part of the problem - maybe the whole thing. The inaddr.arpa address for the IP address of my virtual server resolved to avalon.hartley-web.com a domain I dropped somewhile ago. It should be updating itself to avalon.hartley-consultants.com as we speak. Also - I think I confused both Lee and yourself with the use of 77.96.120.60 as a relay domain I don't actually pass through this virtual server normally - I normally use my ISPs normal mail server. This was just so it was already pre-configured IF I SHOULD NEED IT. -- Alan Chandler http://www.chandlerfamily.org.uk -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/50759b98.1060...@chandlerfamily.org.uk
Re: Exim4 behaviour when long term failure of outgoing address
Alan Chandler a...@chandlerfamily.org.uk wrote: I am using Debian Squeeze on a virtual machine that I lease. It has exim4 (light) version as its mail server. - its name is avalon.hartley-consultants.com However, it looks to me like its trying to send a failure e-mail to me locally somehow. 2012-10-05 07:42:09 1TK1bf-Mx-0C = r...@avalon.hartley-consultants.com U=root P=local S=389 2012-10-05 07:42:09 1TK1bf-Mx-0C ** i...@mynewdomain.com R=dnslookup T=remote_smtp: retry time not reached for any host after a long failure period 2012-10-05 07:42:09 1TK1bt-N0-DT remote host address is the local host: avalon.hartley-consultants.com It's difficult to tell without knowing the precise setup on the machine, but this looks like you've aliased root to the offsite address i...@mynewdomain.com, but then you've got an entry somewhere that tells avalon that it *is* mynewdomain.com. This could be an entry in /etc/hosts, an MX or A record in DNS, or some fancy aliasing somewhere associated with exim itself. Avalon accordingly tries to deliver to info, locally, and finds that this does not exist. Because it's already trying to deliver a bounce message it simply discards the bouncing bounce and aborts. Unfortunately, without knowing what mynewdomain.com really is, I can't run any non-local diagnostics for you. Unless mynewdomain.com really is yours, in which case you've got a configuration problem there because it's not accepting mail. Chris -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/5u89k9x32a@news.roaima.co.uk
Re: Exim4 behaviour when long term failure of outgoing address
Alan Chandler a...@chandlerfamily.org.uk writes: Anyone any idea what is happening - and how I can change the configuration of exim to tell me when mail, like the above, is being refused. It's rather difficult to figure out what's going on without knowing how your exim is configured. Set up a DNS server first, then abandon the unmaintainable splitfile configuration, copy over the example configuration from exims examples and adapt it to your needs as a start. You will then know what is configured and can figure out what happens. My guess is that exim doesn't accept incoming messages for i...@mynewdomain.com (perhaps because mynewdomain.com is not in the domain list which is used to identify domains for which incoming mails are to be delivered on the local host) and considers mynewdomain.com as a remote host. Then it figures out that mynewdomain.com is the same host as avalon.hartley-consultants.com (you didn't say if it accepts incoming mail to be delivered locally for this domain) and cannot deliver the message for reasons I cannot know from the information you provided. I suspect you might have some sort of DNS issue involved here as well. What are the MX entries like? -- Debian testing iad96 brokenarch -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/87sj9tkr80@yun.yagibdah.de