Hello to all, I have a problem with a Debian 8 server installed following "The Perfect Server - Debian 8 Jessie (Apache2, BIND, Dovecot, ISPConfig 3)" I have a customer that have 120+ email accounts and all of them are connected from the same IP address, they use Thunderbird for email client (but I have try also with Outlook 2016 and Mail on Apple) and the can't read/write email because thunderbird still go in timeout. I have try to change mail_max_userip_connections in /etc/dovecot/conf.d/20-imap.conf (actually I have set it to 35000) but w/out any success (I have ofcourse restart dovecot service and also try to restart the server but nothing change). Really I don't know what to do to fix that problem, this is my first experience with Debian, in past I have always use CentOS and I have never had this kind of problem Any help will be very appreciate as usual. Thanks in advance, Alessandro.
Sounds like a connection issue. Are you using fali2ban? If so stop it. Can you do a simple connect test with telnet? Telnet my.mailserver 25 Verify your getting a response back.
Hello webguyz and thanks for your reply. I've try to connect via telnet (both with fail2ban running & stopped) and I connect w/out problem, if can help, I can also login w/out problem to mailboxes using Roundcube and also with Thunderbird, Outlook (any version) Mail and so on if I use an IP different from the one used by the customer (and he have a static IP so it never change), customer can use mailclients w/out problem till there are at least 5 people/mailclients, if someone else turn on his mailclient (and sure they do) than I got this issue. Any other customers on that mailserver can use their mailclients w/out issues also if they have same mailboxes on desktop, smartphones and tablet. Issue is only with that customer that have a lot of mailboxes from the same place and IP.
Maybe take a look some postfix info about simultaneous deliveries? http://www.postfix.org/TUNING_README.html#rope
Are you showing any errors/issues in your /var/log/mail.err or mail.warn files? Also are you saying on TB is having this issue, or any email client from that IP?
Try posting on a Postfix forum? If a problem has seen before it would the best place to try and find the problem.
Thanks webguyz but no, is not an smtp problem, I have problem with IMAP, I can see it from Mail (MACOS debug logs) I haven't any SMTP timeout but I have it on IMAP.
If I try from any email clients I have the issue, but if I use webmail, I have no issues also if I connect all the mailboxes from the same IP but via webmail, on CentOS same hardware I have never had this kind of issues, maybe I'll setup a new server and migrate from Debian 8 to CentOS 7, I have try Debian but maybe is not for me.
Is not a Postfix issue, but a dovecote ones, I have no problem to delivery mail messages, but I have problems to let user connect via mailclients from the same IP address, if I connect the same mailboxes via emailclients (any kind of that) but via different IP addresses I have no issues, issues are only from the same IP address, that's why I have think about mail_max_userip_connections and I have change that value to a big one but with no success.
/var/log/mail.err contains only 2 lines with unexpected command-line argument: reject, but they are of today, one at 10:35 am and one at 4:45 pm, /var/log/mail.warn contains only info about failed login from unknow (someone who try to access mailserver w/out fortune, eh eh) and I have check all IP and no one is the one of my customer (all IP are out of Italy, all my customers are in Italy like me).
