Normally the cpu load comes when you use the history tab, we should go to the legacy history like I mentionned aboveâŠduring the time the lead dev fixes the issue.
If the load comes without consulting the history, please raise your hand
Do you have an idea of when it will be fixed?
I tested nethserver-mail2-* on 2 servers, which were already running mailservers so I switched to mail2:
Both servers were fully updated with nethserver-mail2 2.0.0-1.45
One server with getmail and smarthost is just working without problems, except of:
Iâve gone through test case 2.2:
- The sample file sample-virus-simple2.txt does not exist
- sample-virus-nested.txt goes through and is then detected by my client
On the other server with direct mailing I get an error on screen and in /var/log/messages when I browse to âEMail Addressesâ (MailAccount) in web UI.
Feb 18 23:05:56 nethserver httpd: [64] Cannot use lexical variable $view as a parameter name - File /usr/share/nethesis/NethServer/Module/MailAccount/User.php, line 80
I also have no dkim signature in the text field to enter it in the TXT record. This is the server where the âJunkâ folder is reappearing, see this thread:
First of all, thank you for testing @mrmarkuz!
Youâre right itâs probably a variant invented by me
I noticed it too. Well it seems that test has really strict requirements to run, and we dropped support for it long time ago. So you can ignore the failure - the reason is explained here:
Enhancement #2093: amavisd-new 2.8.0 from EPEL - NethServer 6 - NethServer.org
Iâm fixing the test case!
Did you install the latest package (4 days ago) Rspamd as a new nethserver-mail-filter · Issue #5394 · NethServer/dev · GitHub ?
I cannot see that message in my log: are you using a different PHP version (SCL)? Is it a warning or error? Is the web page correctly displayed or does it contain any error message? Did you increase the log verbosity by some means?
I recall a similar php warning with php7X and scl
Exactly, uninstalling nethserver-php-scl did the trick
- MailAccount working
- dkim signature is working
Thank you for your support! Will it be possible to use PHP7 with nethserver-mail2 ? Itâs needed for moodle for instance.
We must not install PHP7 in httpd-admin instance. I think the correct approach is not replacing the default upstream (PHP5) version with the SCL one. Instead, setting up a php-fpm worker can help in maintaining different PHP versions - like nethserver-nextcloud does.
When I allowed to use php-mod with php7X for the whole server there were no issue but it is no more the case so I should forbbid it now.
Maybe i could think a solution for those use /var/www and want another php version.
That would be nice as I really like the php-scl module but what about the php-fpm approach @davidep suggested? I have to find out how it works in nextcloud but if it would be possible to use other PHP versions without remi SCL it may be the better wayâŠ
When you install nethserver-php-scl you use
Php-fpm for virtualhost
Php-mod for the whole server (virtualhost and /var/www)
The goal could be to use only php-fpm
This the template to activate fpm https://github.com/stephdl/nethserver-php-scl/blob/ns7/root/etc/e-smith/templates/httpd/vhost-extra/15PhpScl_values
DKIM depends on this package, right?
Yes dkim is part of nethserver-mail2-common
I reinstalled the PHP scl module and MailAccount and dkim is still working. Iâll keep you updated if something goes wrong againâŠ
New version of rspamd, maybe we can do a reverse proxy of the webui of rspamd in tbe server-manager, the password issue seems solved. For the history issue I need to check, it is not part of the changelog
after updating the rspamd to new version i receive:
got IO timeout with server rspamd.com(88.99.142.95), after 3 retransmits
formerly it was okay
working smooth on my side, try to restart redis-rspamd and rspamd
systemctl restart rspamd-redis rspamd
as a side note, we agreed with @davidep to release by the nethserver repository the version of rspamd, obviously we need to test them
Same timeouts here:
grep "got IO timeout" /var/log/maillog | wc -l
2199
@hucky @filippo_carletti sorry for my mistakeâŠstill in holiday
the IO error is relative to fuzzy check database of rspamd.com (the main website)
according to this
https://groups.google.com/forum/#!topic/rspamd/OtizdoxmLFs
https://groups.google.com/forum/#!topic/rspamd/5_Zc2Ynm4Dc
it could be a firewall issue (UDP 11335) or a routage problem, please try it, I cannot reproduce it.
PSâŠtime to play with my daughter
i see that the error comes from the controller:
21.2.2018, 11:00:03 controller 17793 csession c2ac55 got IO timeout with server rspamd.com(88.99.142.95), after 3 retransmits
but in any case it seems not to be every time it makes the check.
i will try to figure out more.
PS Good Reason to do a break, wish you a nice time with your daughter
Just added p3scan
module to the mail2 bundle.
New test cases 1.4 and 2.4, added here /cc @quality_team
https://github.com/NethServer/dev/issues/5394#issuecomment-367295232
Updated docs
http://docs.nethserver.org/en/latest/mail2.html#from-pop3-proxy-module
This issue requires QA tests too: