Zextras Forum - Recent Topics

Troubleshooting error messages from IMAP clients attaching via LAN to new carbonio CE server

BISI_II — Tue, 28 Apr 2026 06:28:55 +0000

so - after working on this for a few hours, this has become basically an entry for posterity, in the hope that zextras lasts long enough to make this a searchable resource. I'll also post back here with the results of my flailing. I still want a decent reference, but I'm not holding my breath.

Here is the problem description, in brief.

I am migrating from zimbra FOSS 8.8.15 to carbonio CE. I appear to have hit some kind of connection limit on the new server that I long ago fixed on the zimbra server (although that is a conclusion drawn from the intensive searching I have been doing. I have no notes of the changes, but the fixes are changes to the zimbra defaults). Specifically, we have approx 13 IMAP clients connecting to the same account on either server. When we get to attempting to connect the 12th and 13th clients to the carbonio server, Betterbird (the IMAP client) takes about 90 seconds to timeout and report "unable to connect to .... If the problem persists you may have exceeded the maximum number of connections to this server" (I see no way to include a screenshot, except as a link, which will be useless in very short order).

Is there a command-line reference for the carbonio prov commands and options? The 'carbonio help prov' command gives less-than-useful hints for this particular situation.

In particular, is there a provisioning command for carbonio CE for thie ZimbraImapNumThreads and zimbraImapMaxConnections? This is what I (must have) used for zimbra FOSS 8.8.15 ('cause the defaults recorded in various search hits from years ago for zimbra are both 200), since the current setting for both on the zimbra server is 500.

zmprov ms `zmhostname` zimbraImapNumThreads 500
zmprov ms `zmhostname` zimbraImapMaxConnections 500

This command, after a long wrestling match with claude and gemini, to get zextras to give up its secrets gives me a clue...

carbonio prov gs $(zmhostname) | grep -i imap
  zimbraImapMaxConnections: 200
  zimbraImapNumThreads: 200

The attributes exist! Now, how the bleep do I change them?

zextras@mail:~$ carbonio prov ms $(zmhostname) zimbraImapMaxConnections 500
zextras@mail:~$ carbonio prov ms $(zmhostname) zimbraImapNumThreads 500
carbonio prov gs $(zmhostname) | grep -i zimbraImap
  zimbraImapMaxConnections: 500
  zimbraImapNumThreads: 500

now all I need to do is wait for the bulk email migration to complete so I can restart and see if it truly does have an effect

Troubleshooting error messages from IMAP clients attaching via LAN to new carbonio CE server

BISI_II — Tue, 28 Apr 2026 06:20:37 +0000

Here is the problem description, in brief.

Is there a command-line reference for the carbonio prov commands and options? The 'carbonio help prov' command gives less-than-useful hints for this particular situation.

zmprov ms `zmhostname` zimbraImapNumThreads 500
zmprov ms `zmhostname` zimbraImapMaxConnections 500

This command, after a long wrestling match with claude and gemini, to get zextras to give up its secrets gives me a clue...

carbonio prov gs $(zmhostname) | grep -i imap
  zimbraImapMaxConnections: 200
  zimbraImapNumThreads: 200

The attributes exist! Now, how the bleep do I change them?

zextras@mail:~$ carbonio prov ms $(zmhostname) zimbraImapMaxConnections 500
zextras@mail:~$ carbonio prov ms $(zmhostname) zimbraImapNumThreads 500
carbonio prov gs $(zmhostname) | grep -i zimbraImap
  zimbraImapMaxConnections: 500
  zimbraImapNumThreads: 500

now all I need to do is wait for the bulk email migration to complete so I can restart and see if it truly does have an effect

"For technical discussion, threads, conversation, and solutions please feel free to use the Zextras Forum dedicated to Carbonio CE"

BISI_II — Tue, 28 Apr 2026 05:01:01 +0000

Where, exactly, can I find the forum described in the resources described in the "title page" of this forum?

Would dearly love to find a community discussion of how to do things in CE by people who have done it.

Carbonio CE Migrating from One Server to Another – A Simple Method?

jane — Mon, 27 Apr 2026 09:38:59 +0000

Hello everyone,

I'm currently working with Carbonio CE and I'm looking for the best way to migrate a complete instance from one server to another, simply and reliably.

Context:
Old Ubuntu 22 server: already in production (existing accounts, emails, and configurations)
New Ubuntu 24 server: clean, ready to receive Carbonio
Objective: migrate all data (accounts, emails, configurations, etc.)
My questions:
Is there an official or recommended method for this migration?

Is it better to:
restore from a backup?

or use a synchronization tool?

What are the key steps to avoid data loss?

Are there any sensitive points (DNS, certificates, permissions, proxy…) to monitor?

If anyone has already done this migration, I would appreciate any feedback or a step-by-step guide.

Thank you in advance for your help.

Carbonio CE 26.3 – Ubuntu 24.94 Auth v3 (/zx/login/v3/*) returns 502 from Internet, works only inside LAN (Envoy issue?)

Giovanny-E — Mon, 27 Apr 2026 06:39:41 +0000

Hello,

I am running Carbonio CE 26.3.0 on Ubuntu 24.04 (single-server setup).

Issue:

Mobile apps (Chats / Mobile login) fail from Internet with:

/zx/login/v3/auth/config → HTTP 502 (Connection to upstream refused)

However:

- Web UI works normally

- Mobile apps work when the client is inside the same LAN as the server

- Services are all running and “green”

Technical findings:

- carbonio-user-management is healthy and listening on an internal IP:

127.78.0.5:10000 (confirmed via ss -lntp)

- Envoy sidecar listens on 127.78.0.5:20000

- NGINX routes /zx/* through Envoy by design

Tests performed:

- Direct access to the backend does NOT expose Auth v3:

curl removed link

(no response (no JSON))

This confirms:

- Auth v3 endpoints are NOT exposed by carbonio-user-management directly

- Envoy is REQUIRED to publish /zx/login/v3/*

- There is no viable NGINX bypass/workaround for Auth v3

Observed behavior strongly suggests:

- Envoy does not mark the Auth v3 upstream as healthy in CE 26.3

- This only affects traffic coming from outside the LAN

- Results in consistent 502 errors for mobile apps

Has anyone else confirmed this behavior on CE 26.3 (Ubuntu 24.04)?

Is there a known issue or planned fix regarding Envoy/Auth v3 upstream health?

Thanks,

Giovanny.

failed systemd units on new install of carbonio ce - implications and how do I troubleshoot?

BISI_II — Thu, 23 Apr 2026 21:42:27 +0000

script-based single server installation of carbiono ce 26.03.0 on Ubundtu server 24.04.4

I have a basically unconfigured carbonio server - the password for zextras has been changed, and I've added a test user (to determine where I need to provision storage for the the mail I will be migrating here). In the rehersals/testing, e.g. shutting down all the carbonio services cleanly I have discovered that after some amount of idle time and/or light testing (sending email between two on-host users) that service units (not sure of the correct jargon) have failed:

Get the list of all Carbonio services

systemctl list-unit-files

get a status of the whole systemd

systemctl status
● mail.domain.com
   State: degraded
● mail.doamin.com
   State: degraded
   Units: 556 loaded (incl. loaded aliases)
    Jobs: 0 queued
  Failed: 2 units
   Since: Tue 2026-04-21 23:19:36 PDT; 1 day 15h ago
 systemd: 255.4-1ubuntu8.15

so - what has failed?

~# systemctl list-units --failed
   UNIT                             LOAD   ACTIVE SUB    DESCRIPTION                   
   ● carbonio-docs-editor.service     loaded failed failed Carbonio Docs Server
   ● carbonio-preview-sidecar.service loaded failed failed Carbonio Preview sidecar proxy
 
   Legend: LOAD   → Reflects whether the unit definition was properly loaded.
           ACTIVE → The high-level unit activation state, i.e. generalization of SUB.
           SUB    → The low-level unit activation state, values depend on unit type.
  
  2 loaded units listed.

anybody with any knowledge about what this means for ongoing operations, and what, if anything I can do about it -- please weigh in.

Why, after shutting down all the carbiono CE 26.3 services on ubuntu 24.04.4 does /usr/bin/envoy continue to run (with 1 zombie following it around)?

BISI_II — Wed, 22 Apr 2026 01:55:26 +0000

Shutting down carbonio ce before downing the server.

root@mail1:/opt/zextras# systemctl stop carbonio.target
root@mail1:/opt/zextras# systemctl status carbonio.target
○ carbonio.target - Zextras Carbonio
Loaded: loaded ( removed link ; enabled; preset: enabled)
Active: inactive (dead) since Tue 2026-04-21 18:25:34 PDT; 11s ago
Docs: man:systemd.special(7)

Apr 20 18:13:50 mail1.kibro.com systemd: Reached target carbonio.target - Zextras Carbonio.
Apr 21 18:25:34 mail1.kibro.com systemd: Stopped target carbonio.target - Zextras Carbonio.

top - 18:40:10 up 1 day, 27 min, 1 user, load average: 0.02, 0.11, 0.45
Tasks: 294 total, 1 running, 292 sleeping, 0 stopped, 1 zombie

here is what I am talking about...
as an aside, the whole carbonio mesh thing seems to generate 26-29 zombie
processes at cold boot...

root@mail1:/opt/zextras# ps aux | awk '$8 ~ /^/'
carboni+ 3733 0.0 0.0 0 0 ? Z Apr20 0:00  
root@mail1:/opt/zextras# ps -o ppid= -p 3733
3200
root@mail1:/opt/zextras# ps aux | grep 3200 | grep -v root
carboni+ 3200 0.7 0.0 2398496 62096 ? Ssl Apr20 10:51 /usr/bin/envoy --config-path  removed link  --disable-hot-restart

Can I search the zextras community forum(s) ?

BISI_II — Tue, 21 Apr 2026 23:54:29 +0000

I would really like to be able to search the forums on this site for answers to questions that must have been answered by now. It helps in so many ways, especially for users like me with long experience in the field who are attempting to deploy the product (note that I created the account over 6 years ago).

Unfortunately, clicking on the little magnifying glass has no effect, and I'm guessing that is its intended function, since there is no mouse-over to actually describe it. This is true with Firefox and Chrome, on both Linux and windows

So, is it your site, or is it me?

550-No Such User Here

tanka — Tue, 21 Apr 2026 02:55:10 +0000

Hi,

Is there a way to enable Carbonio to verify the user to the recipients domain? It looks like Carbonio is not doing this. I sent the email from Carbonio webmail so it should have authenticated.

Here is the bounced email I got.

: host nauoi.com said: 550-Verification failed for 550-No Such User Here
550 Sender verify failed (in reply to RCPT TO command)

Really appreciate how I can solve this problem so that if there are very strict, our mail server will be able to comply.

I am using Carbonio CE 24.9.5.

Carbonio is on hq.pub.com.ki

Thanks in advance.

Tanka