It's stuffed inside the username portion of the target URL. We already mangle the target before passing it to duplicity so there wasn't a need for a new field.
Fixes the issue raised in #2200, #2216.
We were using duplicity 0.8.21-ppa202111091602~ubuntu1 from the duplicity PPA probably until June 5, which is when my box automatically updated to 0.8.23-ppa202205151528~ubuntu18.04.1. Starting with that version, two changes broke backups:
* The default s3 backend was changed to boto3. But boto3 depends on the AWS SDK which does not support Ubuntu 18.04, so we can't install it. Instead, we map s3: backup target URLs to the boto+s3 scheme which tells duplicity to use legacy boto. This should be reverted when we can switch to boto3.
* Contrary to the documentation, the s3 target no longer accepts a S3 hostname in the URL. It now reads the bucket from the hostname part of the URL. So we now drop the hostname from our target URL before passing it to duplicity and we pass the endpoint URL in a separate command-line argument. (The boto backend was dropped from duplicity's "uses_netloc" in 74d4cf44b1 (f5a07610d36bd242c3e5b98f8348879a468b866a_37_34), but other changes may be related.)
The change of target URL (due to both changes) seems to also cause duplicity to store cached data in a different directory within $STORAGE_ROOT/backup/cache, so on the next backup it will re-download cached manifest/signature files. Since the cache directory will still hold the prior data which is no longer needed, it might be a good idea to clear out the cache directory to save space. A system status checks message is added about that.
Fixes#2123
By not advertising SMTPUTF8 support at the start, senders may opt to transmit recipient internationalized domain names in IDNA form instead, which will be deliverable.
Incoming mail with internationalized domains was probably working prior to our move to Ubuntu 18.04 when postfix's SMTPUTF8 support became enabled by default.
The previous commit is retained because Mail-in-a-Box users might prefer to keep SMTPUTF8 on for outbound mail, if they are not using internationalized domains for email, in which case the previous commit fixes the 'relay access denied' error even if the emails aren't deliverable.
The /admin/munin routes used the same Authorization: header logic as the other API routes, but they are browsed directly in the browser because they are handled as static pages or as a proxy to a CGI script.
This required users to enter their email username/password for HTTP basic authentication in the standard browser auth prompt, which wasn't ideal (and may leak the password in browser storage). It also stopped working when MFA was enabled for user accounts.
A token is now set in a cookie when visiting /admin/munin which is then checked in the routes that proxy the Munin pages. The cookie's lifetime is kept limited to limit the opportunity for any unknown CSRF attacks via the Munin CGI script.
Port 465 with "implicit" (i.e. always-on) TLS is a more secure approach than port 587 with explicit (i.e. optional and only on with STARTTLS). Although we reject credentials on port 587 without STARTTLS, by that point credentials have already been sent.
All A/AAAA-resolvable domains that don't send or receive mail should have these null records.
This simplifies the handling of domains a bit by handling automatically generated subdomains more like other domains.
This change will force everyone to be logged out of Roundcube since the encryption key and cipher won't match anyone's already-set cookie, but this happens anyway after every Mail-in-a-Box update since we generate a new key each time already.
Fixes#1968.