ZhangHuangbin wrote:

btw, better create a new ticket on iRedMail Store website for EE issues.
https://store.iredmail.org/

One of my two servers also had the lock issue, and I had to keep trying until it worked.

Doesn't it cost 150 for a ticket?

Creating ticket on iRedMail Store website is free.

btw, better create a new ticket on iRedMail Store website for EE issues.
https://store.iredmail.org/

One of my two servers also had the lock issue, and I had to keep trying until it worked.

Doesn't it cost 150 for a ticket?

/opt/iredmail/custom/bind/** kr,

- Does restarting apparmor service fix the permission issue?

wget installer, complete with GUI as directed.
Ubuntu 24.04
LDAP BE
NGINX
iRedAdminPro mgmt

Couple of problems with the trouble free upgrade today on Ubuntu 24.04. Probably easily fixed. One apt lock collision and one bind-app armor block.

Thanks,
Tim

You need to toggle OFF option "Disable spam scanning" to enable the scanning.
Will improve it in future release shortly.

Thanks.

I manage SPAM/AntiVIRUS before my EE until now  (disabled spam and virus scanning).

I decided to add another step of protection for viruses and enabled it.  Is this setting correct or do I have to enable SPAM scanning in the 1st paragraph?

ERROR: directory doesn't exist or incorrect permission (check parent directories also): /var/vmail/mlmmj.

it is set to vmail:vmail

this is the service:

mlmmjadmin.service - RESTful API server used to manage mlmmj mailing list manager
     Loaded: loaded (/usr/lib/systemd/system/mlmmjadmin.service; enabled; preset: enabled)
     Active: failed (Result: exit-code) since Mon 2026-08-31 13:25:28 CDT; 1s ago
   Duration: 100ms
    Process: 52786 ExecStartPre=/bin/mkdir -p /var/run/mlmmjadmin (code=exited, status=0/SUCCESS)
    Process: 52787 ExecStartPre=/bin/chown mlmmj:mlmmj /var/run/mlmmjadmin (code=exited, status=0/SUCCESS)
    Process: 52788 ExecStartPre=/bin/chmod 0755 /var/run/mlmmjadmin (code=exited, status=0/SUCCESS)
    Process: 52789 ExecStart=/usr/bin/uwsgi --ini /opt/mlmmjadmin/rc_scripts/uwsgi/debian.ini --pidfile /var/run/mlmmjadmin/mlmmjadmin.pid (code=exited, status=255/EXCEPTION)
    Process: 52795 ExecStopPost=/bin/rm -rf /var/run/mlmmjadmin (code=exited, status=0/SUCCESS)
   Main PID: 52789 (code=exited, status=255/EXCEPTION)
        CPU: 109ms

Error: ConnectionError(MaxRetryError("HTTPConnectionPool(host='127.0.0.1', port=7790): Max retries exceeded with url: /api/subscriber/eddiegarner@progressiveelectricalservices.com/subscribed?query_all_lists=no&email_only=yes (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7ae08f37ad20>: Failed to establish a new connection: [Errno 111] Connection refused'))"))

I have tried restarting mlmmjadmin.service, when status is checked it is failed to start.

What am I missing?

Looks like someplace in between iredapd versions it now wants python-multipart ... there is not a distro multipart package for RHEL8 as best I can tell.

So, while I can probably "pip install" this someplace and modify the PYTHONPATH in the systemd service file, it might be good for EL8 to simply include this in iredapd with a "pip install --prefix=/opt/iredapd..." if we're running on EL8.

thanks!

Hi,
I'm trying to set a fractional spam score via the RESTful API (PUT /api/spampolicy/user/<mail>, same for domain and global), e.g.:
spam_score=4.5
According to the docs docs.iredmail.org/iredadmin-pro.restful.api.html#spam-policy , the spam_score parameter is described just as "spam score", with the sample using an integer (spam_score=4), but it's not explicitly stated that decimals are disallowed.

I checked the source and found the relevant code in libs/amavisd/spampolicy.py.

str.isdigit() returns False for any string containing a decimal point, so spam_score=4.5 silently fails the if check — the block is skipped entirely, no error is returned, and spam_tag2_level/spam_kill_level are simply left unset. The API still responds with {"_success": true}, which is misleading since the value wasn't actually applied.

Interestingly, the code itself calls float(v) right after the isdigit() check, which suggests decimal support was intended but blocked by the wrong guard condition (should probably be a try/except around float(v) directly, rather than isdigit()).

Questions:

1. Is integer-only spam_score an intentional limitation, or is this a bug in the input validation?
2. Through the iRedAdmin-Pro web GUI, decimal spam score values are accepted and correctly applied/stored — so this inconsistency seems specific to the RESTful API input handling. Can you confirm the API is meant to behave the same way as the GUI here?
3. If it's a bug, could it be fixed in an upcoming release? Happy to test a patch if provided.