Tag Info

Hot answers tagged

9

The general answer is - NO. The reason is that while you CAN put your server to receive on that port, the other servers i the world will send their email to port 25, and you can not change that. This is part of the SMTP specficiations, and while you are free to configure your server different - that is of no practical use unless the rest of the world cares ...


6

In My Humble Opinion.. It shouldn't. Managing a mail-server, properly, is a massive ball-ache. We have 40 odd users, and use GAFYD (Google Apps For Your Domain). It works perfectly. We don't need Exchange or any of that Microsoft Bloatware. It's nice not to have to worry about spam, or anything like greylisting. I'm very happy GAFYD. I can also highly ...


6

Resizing an iSCSI LUN on the fly on a NetApp is no big deal, we do it all the time. The only gotcha is in the level of support of your file system for resizing a block device on the fly. That said the ease of using snapshots and WAFL log file system are two big benefits for NFS on a NetApp. While I don't deal with Zimbra or any other mail servers in my line ...


6

Having a properly configured SPF record can alleviate a little bit of that pain. It will prevent someone from sending email claiming it is from your domain.


5

Bulk email sending is a specialist’s role these days, that's why you're getting the advice you are. The big mail vendors like Mandril and Sendgrid maintain farms of machines not because of scale, but because of IP reputation. One of the most effective spam fighting techniques available to us is a database that tracks how well behaved IP addresses are with ...


4

Use a dedicated line to define your FQDN. 192.168.10.9 mail.mydomain.com mail Remove it from the 127.0.0.1 line.


4

Let clear some side-question before proceed into main question: Will this files be archived while update of zimbra? Or they will be overwritten while updating? Yes, Zimbra will overwrite master.cf.in and all of zimbra predefined configuration templates. So what I've got here - Seems like Zimbra 8.6.0 is not loading this file, since there is no ...


4

If you do implement this, it will likely do more harm than good. A lot of people do this when Bcc'ing a large group of people. They put themselves in the To: field and Bcc: everyone else. You'll just end up up with a lot of false positives. Just ensure that you've classified this and other emails like it as spam in SA and let the heuristics do their thing.


3

I had the same problem trying to configure an account to act as a catch all address for multiple domains and found this valuable forum entry: catch all not working: Multi-value syntax The command zmprov modifyAccount catchall@mydomain.com zimbraMailCatchAllAddress @mydomain.com sets the zimbraMailCatchAllAddress attribute for account ...


3

You don't need a dedicated DNS Server for zimbra, you can user your existing network DNS, add zimbra entries there and make your zimbra server use that DNS server. As a workaround to that DNS issue, it happens to me also, when no DNS server is available, to use /etc/hosts file to resolve zimbra servers hostnames internally.


3

Tighten the config, make sure it's not relaying, and clean everything up. Then contact the blacklist sites and have them retest. That's really all you can do, find who's blacklisting you after cleaning things up and contacting them to re-evaluate your site and be taken off the lists.


3

I don't see anything wrong with the HR plan. It makes sense to me to keep the emails separated by role. If the additional licensing cost is approved then my opinion would be to go with it. You have a point about a user accessing email sent to one role from the other role, but I think the point is to take appropriate steps to keep the roles distinct and ...


3

You have to use either different IPs or different ports. SNI (deciding which SSL certificate based on the name the other person is trying to reach) isn't supported on enough browsers yet for name-based SSL hosting to work.


3

I can't really test this because I don't have a test setup of Zimbra to work with, but I'd suspect adding the zimbra user to AllowUsers and setting the ssh port back to 22 should be sufficient. If you're nervous about exposing ssh on a standard port, it should be possible to configure sshd to listen on your alternative port as well as port 22, and use ...


3

There are a number of reasons why you might want (or need) to have a local mail server. Here are a few: Control: your mail server is run your way, with no unexpected changes. Security: even if you are setup so that not-specifically-encrypted mail from your clients is transported securely rather than in plain text it will probably be stored plain on the ...


3

The Zimbra resource requirements are relatively lightweight compared to the needs for an Exchange 2010 system. I recently converted my firm's mail solution from a hosted Zimbra setup to an internal Exchange 2010 system. We are 100 users with ~570 public-facing mail addresses (from users, complex distribution groups and mail-enabled public folders). The ...


3

I use to run about 5-6 [any] to Zimbra migration per month, and I use ImapSync (latest 'paid' version, but should be no problem if you are using an older one) with delegated access 90% of the times. Have you tried the --authuser1 and --authuser2 options? The line you use could be modified as: imapsync --nosyncacls --syncinternaldates --host1 ...


3

The imapsync --authuser1 option explained by Jazzy Pierre may solve your issue. Have a try. (I wrote imapsync).


3

I would use zmmboxsearch on the command line is a really easy way to dump messages. So if you wanted to dump all the messages in the inbox, as the zimbra user do: zmmboxsearch -d /tmp/user/ -m user@domain.com -l 10000 -q "in:inbox" This will dump all the raw messages that match the query for user in /tmp/user. The "-l" is the limit, by default it only ...


3

Something like (uid=%u) should work, or given your ldapsearch example, perhaps (&(objectClass=JammMailAlias)(mail=%u@domain.com)) Zimbra replaces the %u with the username that is attempting to authenticate, and then does a search/bind as that user to authenticate. Edit: In your setup, you should be able to use (mail=%u@domain.com) as your search ...


3

I suppose it could have something to do with the machine "not being able to find itself". I make a practice of putting the machine's own IP address in /etc/hosts, so that whether a particular service is bound to 127.0.0.1 XOR to the external address, it will be able to connect to that service. I agree, you have provided very little info, and a very broad ...


3

But is it possible to extract the existing accounts from Zimbra including authentication credentials, accounts and aliases, distribution lists, etc -- to run the same servers a la carte, without the Zimbra "suite" involved? YES First task that you needs was parsing zimbra LDAP data extracted with this command (taken from this page) ...


3

Have you considered using a mail journaling solution? Searching across that large a dataset seems like it would be difficult with what you're proposing. My preferred method is to use MailArchiva in its free or commercial version. It works with Exchange, Sendmail, Zimbra, etc. and is powered by the Apache Lucene search engine. This would leave you free to ...


3

As others pointed out your SMTP relay has probably changed something. Looks like they chagned their advertised authentication mechanisms. I don't know if swaks tells you which authentication mechanism it used in it's output, but you could try specifying an auth-type and see if you get through. The parameter for digest-md5 would be: swaks -s smtp.strato.de ...


3

If you are using the etoken (pki-usb-authentication) you will probably do Client Certificate authentication in the browser. This would require to reconfigure the Apache server to require a client certificate. And here is the problem. You can not determine which user is about to authenticate, before the user/browser provided the client certificate. Thus users ...


3

Of course you can do this. Subdomains are just separate VirtualHost configurations. The names don't matter. i.e., to Apache, mail.domain.com and mail.anotherdomain.com are as different as mail.domain.com and mail2.domain.com. e.g.: <VirtualHost 192.168.1.1:80> ServerName mail.mysite.com DocumentRoot /home/mysite.com/sites/mail/public_html ... ...


2

root@yourdomain.com is default account on many Linux distributions, and just should exists. On that mail you are getting for example information about failed cron jobs (jobs that write to stderr) and other quite important things. What I suggest: create alias to Your main administrator account root@yourdomain.com in Zimbra admin panel maybe creating ...


2

Follow this link: http://www.maxxer.it/2010/linux/set-apache2-to-proxy-zimbra/ I've successfully proxied Zimbra using Apache2 for long time. These commands work on Debian/Ubuntu servers. At first, enable apache2′s modules: a2enmod proxy a2enmod proxy_html a2enmod proxy_http Make sure the use of mod_proxy is allowed, by changing ...


2

You may want to check that the users in the @spam_lovers_maps correspond to the user who needs to receive any and all spam, and that messages to him do not get caught in the $final_spam_destiny = D_DISCARD; trap. You may also want to ensure that the user does not have any .procmail files, or similar delivery time scripts that may discard the mail after ...


2

Don't do this. Zimbra is a complicated beast with lot of components and it should run on a dedicated server or VM. For the specific problem: Obviously, you noticed that there is a conflict between Zimbras own web server and the one you installed. If you insist to continue with this, you would have to change ports for the Zimbra web server installation and ...



Only top voted, non community-wiki answers of a minimum length are eligible