[Users] January 19, 2021 Zeta Alliance Conference Call Summary

Noah Price noah-zeta at prxy.com
Sat Jan 23 00:21:15 CET 2021


Thanks Randy, very helpful summary as always! 

To provide some more info for anyone else who is interested, it looks like Cine's suggestion sender_dependent_relayhost_maps is the right direction. That led me to this Zimbra wiki post with more information: [ https://wiki.zimbra.com/wiki/Relay_per_Domain | https://wiki.zimbra.com/wiki/Relay_per_Domain ] 

I have not tested extensively yet, but plan to test and implement this soon. 

Noah 
prxy, Inc. 

> From: "Randy Leiker" <randy at skywaynetworks.com>
> To: "users" <users at lists.zetalliance.org>
> Sent: Friday, January 22, 2021 11:52:37 AM
> Subject: [Users] January 19, 2021 Zeta Alliance Conference Call Summary

> Hello Zeta Alliance Community,

> Here is a summary of this week’s conference call. A few brief reminders:

>    * Conference calls are every Tuesday and open to all using either the
>    FreeConferenceCall.com VoIP app or via a dial-in number: [
>    https://www.freeconferencecall.com/wall/zetalliance |
>     https://www.freeconferencecall.com/wall/zetalliance ]
>    * Each week’s call agenda can be found at:
>     https://drive.google.com/drive/folders/1xDyBJFjnfZYxuXJHiDzsXjjMuGGtIl7J
>     * A copy of each week’s summary is also posted to the Zimbra Forums:
>        * All Prior Months: [ https://forums.zimbra.org/viewforum.php?f=9 |
>         https://forums.zimbra.org/viewforum.php?f=9 ]
>        * December 2020 : [ https://forums.zimbra.org/viewtopic.php?f=9&t=69008 |
>         https://forums.zimbra.org/viewtopic.php?f=9&t=69008 ]
>        * January 2021 : [ https://forums.zimbra.org/viewtopic.php?f=9&t=69121 |
>         https://forums.zimbra.org/viewtopic.php?f=9&t=69121 ]
>     * Constructive feedback on these call summaries is always welcome.

> January 19 , 2021

> Setting A Maximum Message Size Per Domain
> Noah P. said that he is looking for a means to enforce a maximum email size
> limit per Zimbra domain. He explained that he has two customers that want to
> keep email messages within their organizations smaller than they are now, and
> Noah does not wish to apply the lower size limit to all other customer domains
> on his Zimbra servers, but rather just to those two customer’s domains. Cine
> suggested that CBPolicyd ( [ https://wiki.zimbra.com/wiki/How-to_for_cbpolicyd
> | https://wiki.zimbra.com/wiki/How-to_for_cbpolicyd ] ) might be of help, and
> said that he knows CBPolicyd can set a maximum outbound message size limit, but
> was not certain if it was possible to set an inbound message size limit too.
> Mark S. suggested that Noah could setup a new Zimbra MTA server specifically
> for the customers requesting the lower message size limit, then route all of
> the email for those two customers through the new MTA server. Matthew F. said
> that writing a Postfix Milter to filter based on message size might work, but
> the downside to this approach would be that messages that are too large would
> bounce to the sender only after Zimbra initially accepts the messages for
> delivery. Matthew also pointed out that in doing a quick Telnet connection test
> to his Zimbra MTA server, he noticed that Postfix advertises the maximum
> message size accepted immediately after the HELO/EHLO command is sent,
> suggesting that it may only be practical to have a server-wide message size
> limit set, as compared to a per domain setting, assuming that all of the users
> are using the same Zimbra MTA server.

> Resetting Mailbox Passwords When Two-Factor Authentication Is Enabled
> John E. shared an update on an issue discussed in an earlier Zeta Alliance call
> related to difficulties encountered when attempting to use the Zimbra password
> reset feature for a mailbox with two-factor authentication enabled. He said
> that this has been acknowledged by Zimbra as a bug (ZCS-10191; status updates
> are not available via the Zimbra Support Portal) and is anticipated to be fixed
> in an upcoming patch that should be released soon. Noah P. asked if the bug fix
> will only apply to the Modern UI, the Classic UI, or both in Zimbra 9. John E.
> said that the bug fix is associated with the server back-end, so he thinks it
> should apply to both the Modern and Classic UIs.

> Using Shared Calendars Containing The Same Calendar Event UIDs
> From the Zeta Alliance mailing list, Frederic N. said that he has Zimbra users
> who have multiple shared calendars, and they often add the same calendar events
> to these multiple shared calendars with the same UID number. His users report
> that when their email clients are configured to synchronize each of these
> shared calendars, they expect the calendar events with the same UID number to
> appear for each shared calendar, but the actual behavior observed is that the
> calendar events appear only once on a single shared calendar – usually the
> shared calendar that synchronizes first. He wondered if this occurs because of
> the de-duplication feature in Zimbra that also avoids repeatedly delivering
> email messages with the same message ID number. Cine said that Frederic did not
> mention which email clients are in use by the users, but generally it is up to
> each email client to decide on which calendar events they will download, rather
> than the Zimbra server. He did not think that the de-duplication feature in
> Zimbra was involved, as that feature only applies to email deliveries. John H.
> added that most email clients will consider calendar events with the same UID
> to be the same calendar event, thereby causing the event to be displayed only
> once within the email client, even if it appears on multiple shared calendars
> on the server-side. He said that for ActiveSync clients, it is up to the client
> to decide on how to display calendar events with the same UID, as either a
> single event, or multiple events.

> Non-Responsive Mailboxd Process and AWS Storage Buckets
> Mark S. shared that he recently experienced an issue with one of his Zimbra
> mailbox servers where none of the users could access their mailboxes, but the
> mailboxd processes and its dependencies were running with a load of 120%. After
> a restart of the mailboxd and related services, users still could not access
> their mailboxes. With Zimbra Support’s assistance, the issue was pinpointed to
> a problem with mailboxd accessing the AWS storage buckets containing his Zimbra
> server’s disk volumes. He said that when using AWS storage buckets as Zimbra
> volumes, he used to be able to create a primary/secondary volume without also
> needing to create a logical representation in Zextras of the external AWS S3
> storage bucket. But, he discovered that now this is required. Cine said that
> about one year ago, the Zimbra Administration UI changed to add support for
> some updates to the supported volume types, but the Zextras back-end CLI did
> not. He said that it should now no longer be possible to create a Zimbra volume
> without a corresponding AWS S3 bucket. When using AWS buckets for Zimbra
> volumes, Cine recommended running these CLI commands:

>     * “zxsuite hsm getAllVolumes” to get the bucketConfigurationId value, then run:
>    * “zxsuite hsm listBuckets” to see the actual bucketConfigurationId for your
>     actual buckets.
>    * If your primary or secondary storage volume references a non-existent bucket
>     ID, you may need to run “zxsuite hsm doUpdateVolume …” to fix.

> Using Smart Hosts In Zimbra To Manage Outbound Email Routing
> Noah P. re-visited a topic from the November 10th Zeta Alliance call ( [
> https://forums.zimbra.org/viewtopic.php?f=9&t=68942#p299710 |
> https://forums.zimbra.org/viewtopic.php?f=9&t=68942#p299710 ] ) related to
> controlling SMTP mail flow in Zimbra per domain. He said he had been continuing
> to experiment with this capability by working with using the Zimbra Smart Hosts
> feature to attempt to direct mail per domain through specific Zimbra MTAs. Mark
> S. suggested that Noah should use the “zimbraSmtpHostname” setting per domain
> for this purpose, but Noah said this setting only works for users who access
> their mailboxes with the SOAP clients, such as the Zimbra Connector for Outlook
> and ActiveSync. Noah has observed that the Zimbra Web Client, SMTP (IMAP/POP)
> clients, mailbox forwarders, and distribution lists still continue to use the
> Zimbra global MTA setting, ignoring the per domain “zimbraSmtpHostname”
> setting. Cine suggested using Postfix’s “sender_dependent_relayhost_maps”
> attribute in addition to the “zimbraSmtpHostname” setting to help ensure
> consistency for each user’s email client use case. John H. suggested that Noah
> should also create a Zimbra request for enhancement to make this capability
> part of Zimbra’s official features.

> “Host or domain name not found” Errors During Zimbra Server Migration
> Matthew F. said that he is working on a customer migration from an old to a new
> Zimbra server version. He said that after a mailbox is moved from the old to
> the new server, the LMTP server address for the mailbox, that Zimbra uses for
> mail delivery, is updated to the new server’s fully qualified domain name
> (FQDN) as expected, but email deliveries sent from mailboxes on the old server
> to the mailbox on the new server result in “status=bounced (Host or domain name
> not found. Name service error for name=xxxx type=A: Host not found)” error
> messages in mailbox.log. He found that after moving a mailbox to the new Zimbra
> server, and changing the LMTP FQDN for a mailbox to an IP address, this then
> allows for email to be delivered normally. Noah P. wondered if Zimbra may be
> checking the server’s local /etc/hosts file before using the DNS resolver
> configured for the server, and the /etc/hosts file contains an incorrect entry
> for the FQDN of the new server. Mark S. said that he worked with a customer
> that had a multi-homed Zimbra server, with one network interface assigned to a
> public IP address, and another assigned to a VPN. He found it very tricky to
> get mail routed correctly with this setup and suggested using Postfix’s
> “lmtp=native” setting. He said that without this setting, Postfix avoided using
> the /etc/hosts file for look-ups and always used the DNS resolver configured
> for the server. Cine suggested that Matthew take a look at using dnsmasq as it
> makes these types of scenarios much easier. Mark S. suggested taking a look at
> his dnsmasq how-to ( [
> https://www.missioncriticalemail.com/2018/04/12/zimbra-dnsmasq-configuration-guide/
> |
> https://www.missioncriticalemail.com/2018/04/12/zimbra-dnsmasq-configuration-guide/
> ] ) and said that dnsmasq is beneficial since it only answers DNS queries for
> records it is specifically configured to answer, and passes all other queries
> through to the server’s regular DNS resolver. Noah P. also suggested the Zimbra
> command: “zmlocalconfig -e postfix_lmtp_host_lookup=native” which Matthew said
> he currently has enabled. Cine added that using the /etc/nsswitch.conf file to
> set the order in which DNS resolvers are used may help, such as this
> configuration value: “hosts: files dns”. Mark S. said that this page ( [
> https://wiki.zimbra.com/wiki/Postconf_keys |
> https://wiki.zimbra.com/wiki/Postconf_keys ] ) maps Postfix configuration
> settings to Zimbra-equivalent settings may be of help too. Noah P. said that
> Matthew might be looking for this Zimbra setting too:
> “zimbraMtaLmtpHostLookup”, which can be described from the CLI using: “zmprov
> desc -a zimbraMtaLmtpHostLookup”. Matthew F. said he currently has this setting
> configured: “zimbraMtaLmtpHostLookup: dns”.

> Randy Leiker ( randy at skywaynetworks.com )
> Skyway Networks, LLC
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.zetalliance.org/pipermail/users_lists.zetalliance.org/attachments/20210122/df33dd35/attachment.html>


More information about the Users mailing list