MFA (Multi Factor Authentication), SSO, and Cyrus

2020-02-18 Thread John Wade
Hi Cyrus Users, We are currently using Cyrus IMAP with Roundcube webmail, and are looking to implement both SAML or CAS Single Sign-on and Multifactor Authentication (MFA) for all applications.   Currently Cyrus users authenticate back to Active Directory via SASL ldap_auth and this remains

Re: Fwd: Help putting cyrus on Docker

2020-02-18 Thread Fabio Montefuscolo
Hi Nic! I'm super happy with your response! Honestly, I added that while I was getting 550 from lmtp, but that probably should be removed. But lmtp, do you think I can remove that line from the master? Actually, that line came turned on by default on CentOS 8, that is the base image I'm using. I

Re: Fwd: Help putting cyrus on Docker

2020-02-18 Thread Nic Bernstein
Fabio, Very interesting stuff.  I would encourage you to add Readme.md to your repository, containing at least what you've summarized here. This would make for a nice addition to the Cyrus documentation, but would need a bit more explanation on your part, first, so whomever prepares the docs

Re: Fwd: Help putting cyrus on Docker

2020-02-18 Thread Fabio Montefuscolo
Hello! I finally got the basic stuff working on Docker and deployed through Docker swarm. That is what I learnt (or I think I learnt) * a basic mail solution having Cyrus needs 4 containers (imapd, saslauthd, rsyslog and postfix) * sharing rsyslog socket on all containers is needed to have logs

Re: cyrus murder environment and http proxy?

2020-02-18 Thread Jean-Christophe Delaye
On 2020-02-17 12:17, Khalid Mehmood Khan via Info-cyrus wrote: > Hi! > > Does http proxy work in a cyrus murder environment? I keep getting ""PUT > /dav/addressbooks/user/testuser/6ecc0a66-74a1412-1581679269354-210887/ef37977290d43113e709f68848aa9a8a5ff8a24f.vcf > HTTP/1.1" (if-none-match=*) =>

Re: Sieve admin account doesn't work anymore

2020-02-18 Thread Rainer Ruprechtsberger
Hello, > You're trying to use "proxy authentication". That only works with some > SASL mechanisms, not all of them. Yes, I know, we use "PLAIN" which does support that. > Perhaps try the same with imtest and > observe what mechanism is used? imtest is much more verbose than sieveshell. >