On 01/17/2015 10:41 AM, Istvan Gabor wrote:
Hello:
I would like to set up a local email server. I have a linux server which is connected to the internet (WAN) and to a local network (LAN). I want the server to fetch messages from different mailboxes provided by my ISP, and store (and backup) those messages. LAN clients would fetch messages from this linux server and not directly from my ISP.
So far, so good :-) I've run this for a long, long while. Basically it uses fetchmail to read in and and an IMAP server to let local LAN clients access the store. Refinements include using spam/AV checking on the input, filtering of messages into different boxes and more. The big steps is to integrate with Postfix You can find many How-To articles for this on the web. Look/Google for these terms fetchmail, imap, dovecot, spamassassin, procmail, postfix PLEASE NOTE: the above is based you saying "I want the server to fetch messages from different mailboxes provided by my ISP". If you want to have a published gateway of your own domain and and a specific SMTP listener then that is different.
Messages from LAN clients would be sent (and backed up) through/by the server using my ISP's SMTP server. All email traffic (all incoming and outgoing) messages should have a backup which can not be modified by the clients.
You can do this by running something like Postfix or Exim on your server. Clients on the LAN direct outgoing mail to the server. Perhaps your firewall should prevent other than the server sending email out. However that won't stop people using WebMail. If you have something like Postfix running, you can also have fetchmail bringing in the mail and routing though Postfix as the MDA for whatever processing/archiving. I'm hesitant to archive incoming mail unless you are dealing with some legal requirement. SPAM in one form or another, be it true "unsolicited commercial" or just notifications that are read-once and throw away, or simply things that are going to be archived elsewhere such as this list, are all not worth archiving obsessively.
Do I have to set up a mail transfer agent on the server, or use fetchmail to receive the messages, or both? Please give me a few hints how to proceed or what the options are. Links to tutorials and manuals are also welcome.
There's just so much out there on how to do this and minor variations. The "doing" often gives you a better idea of what you actually want and how complex you want it to be. There are many, many micro-decisions along the way, such as the decision to use mbox or maildir formats. All the programs, dovecot, procmail, fetchmail, spamassassin, etc, have many configuration options, and even deciding "I don't need that, ignore it" can be overwhelming. Start small and build. oh, and learn a lot of acronyms! Yes there are canned solutions out there in the How-To, but dropping in the deep end might make debugging horrendous since there's no incremental learning and you won't know what "works so far". -- /"\ \ / ASCII Ribbon Campaign X Against HTML Mail / \ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org