Commit Graph

644 Commits

Author SHA1 Message Date
30c79faff2 fix problem with spf where we would generate errors about too many void lookups
the problem was that we only looked up either the ipv4 or ipv6 address when
evaluating spf directives, depending on the incoming smtp connection. for
example, for spf directive "a", we would lookup the requested domain. if that
domain has an ipv4 address but no ipv6 address, and the incoming connection is
ipv6, we would count a void lookup. but we shouldn't: there is a record for
that name, it just doesn't match the address (family).
2023-03-03 11:45:19 +01:00
1073ca2795 in readme, add link to docker hub, and add your changes will be under mit license too 2023-03-02 20:22:26 +01:00
286411cf82 fix formatting of multiple ptr host names in admin dns check page
from mteege, thanks!
2023-03-02 20:07:02 +01:00
6abee87aa3 improve webserver, add domain redirects (aliases), add tests and admin page ui to manage the config
- make builtin http handlers serve on specific domains, such as for mta-sts, so
  e.g. /.well-known/mta-sts.txt isn't served on all domains.
- add logging of a few more fields in access logging.
- small tweaks/bug fixes in webserver request handling.
- add config option for redirecting entire domains to another (common enough).
- split httpserver metric into two: one for duration until writing header (i.e.
  performance of server), another for duration until full response is sent to
  client (i.e. performance as perceived by users).
- add admin ui, a new page for managing the configs. after making changes
  and hitting "save", the changes take effect immediately. the page itself
  doesn't look very well-designed (many input fields, makes it look messy). i
  have an idea to improve it (explained in admin.html as todo) by making the
  layout look just like the config file. not urgent though.

i've already changed my websites/webapps over.

the idea of adding a webserver is to take away a (the) reason for folks to want
to complicate their mox setup by running an other webserver on the same machine.
i think the current webserver implementation can already serve most common use
cases. with a few more tweaks (feedback needed!) we should be able to get to 95%
of the use cases. the reverse proxy can take care of the remaining 5%.
nevertheless, a next step is still to change the quickstart to make it easier
for folks to run with an existing webserver, with existing tls certs/keys.
that's how this relates to issue #5.
2023-03-02 18:15:54 +01:00
6706c5c84a add basic webserver that can do most of what i need
- serve static files, serving index.html or optionally listings for directories
- redirects
- reverse-proxy, forwarding requests to a backend

these are configurable through the config file. a domain and path regexp have to
be configured. path prefixes can be stripped.  configured domains are added to
the autotls allowlist, so acme automatically fetches certificates for them.

all webserver requests now have (access) logging, metrics, rate limiting.
on http errors, the error message prints an encrypted cid for relating with log files.

this also adds a new mechanism for example config files.
2023-02-28 22:19:24 +01:00
fbfbd97947 connecction -> connection
from mteege
2023-02-28 20:43:31 +01:00
994b5b9a0a fix serving mta-sts policy files
i broke it 3 days ago when adding support for serving it through external reverse proxy.

report by mteege, thanks!
2023-02-28 20:40:52 +01:00
c9190337e6 another rfc (we don't implement it) 2023-02-27 22:35:07 +01:00
d670e68745 make error message about listeners with a missing tls config more helpful for users
by pointing out which section is causing the need for a tls config.
2023-02-27 21:42:27 +01:00
da3fefd42e explain to user they should pick a random, unguessable password
and help with a button to generate one.
2023-02-27 21:29:27 +01:00
aed8360002 more permissions in mox.service to ease with upgrade to v0.0.2
- CAP_FSETID looks needed to make accessing setgid dirs work.
- CAP_DAC_OVERRIDE CAP_DAC_READ_SEARCH are needed to access the existing config/mox.conf.
- CAP_FOWNER seems needed to chmod the config/ and data/ files.
- RestrictSUIDSGID=yes has to be off for chmod to work
2023-02-27 15:48:37 +01:00
cfa6554d03 move sieve up to roadmap, and add jmap there as well, and add questions about staying up to date and learning more about email 2023-02-27 15:04:53 +01:00
e2e2c71212 minor tweaks to suggested dns records 2023-02-27 15:04:32 +01:00
67c97c6499 on admin page, show warning when user hasn't enabled "check updates"
and send Accept-header when fetching changelog, so we can show an html version
instead of json when requested by regular browsers.
2023-02-27 15:03:37 +01:00
d3590caa2e do not require a user "mox" to be present for tests
uid 1000 doesn't actually have to be present (except for the special imaptest),
but this prevents looking up the user named mox.
2023-02-27 14:54:16 +01:00
9e280221b6 merge start.go into serve.go 2023-02-27 14:12:58 +01:00
44a3f9b1bc in imapserver, do not advertise STARTTLS if TLS isn't configured 2023-02-27 14:10:43 +01:00
0639c396b9 add CGO_ENABLED=0 in go install command
because users may run this on their local machine, then copy the binary the
their server. there is a good chance that binary won't start on linux/amd64 when
compiled with cgo due to missing glibc symbols.
2023-02-27 14:05:10 +01:00
259729e03d use full docker hub path in docker-compose.yml 2023-02-27 14:04:45 +01:00
e20677cfd6 add missing capability CAP_FSETID to mox.service
without it, process cannot create setgid directories.
2023-02-27 14:04:20 +01:00
f3f2c6f8ea fix cross-compiled docker images
binaries for linux/amd64 were build for each target platform. apparently the
--platform in the first-stage overrides the $TARGET* variables of the build.
docker behaviour always manages to surprise me...
2023-02-27 13:46:29 +01:00
92e018e463 change mox to start as root, bind to network sockets, then drop to regular unprivileged mox user
makes it easier to run on bsd's, where you cannot (easily?) let non-root users
bind to ports <1024. starting as root also paves the way for future improvements
with privilege separation.

unfortunately, this requires changes to how you start mox. though mox will help
by automatically fix up dir/file permissions/ownership.

if you start mox from the systemd unit file, you should update it so it starts
as root and adds a few additional capabilities:

        # first update the mox binary, then, as root:
        ./mox config printservice >mox.service
        systemctl daemon-reload
        systemctl restart mox
        journalctl -f -u mox &
        # you should see mox start up, with messages about fixing permissions on dirs/files.

if you used the recommended config/ and data/ directory, in a directory just for
mox, and with the mox user called "mox", this should be enough.

if you don't want mox to modify dir/file permissions, set "NoFixPermissions:
true" in mox.conf.

if you named the mox user something else than mox, e.g. "_mox", add "User: _mox"
to mox.conf.

if you created a shared service user as originally suggested, you may want to
get rid of that as it is no longer useful and may get in the way. e.g. if you
had /home/service/mox with a "service" user, that service user can no longer
access any files: only mox and root can.

this also adds scripts for building mox docker images for alpine-supported
platforms.

the "restart" subcommand has been removed. it wasn't all that useful and got in
the way.

and another change: when adding a domain while mtasts isn't enabled, don't add
the per-domain mtasts config, as it would cause failure to add the domain.

based on report from setting up mox on openbsd from mteege.
and based on issue #3. thanks for the feedback!
2023-02-27 12:19:55 +01:00
eda907fc86 better error message if import fails, only stack traces for unexpected panics (i.e. not the special sential panic value) 2023-02-26 22:25:57 +01:00
9b3a170cc2 fix nil pointer deref when importing a message that the junkfilter could not parse (e.g. malformed header)
import does its own batched junkfilter training, so the deliver function
doesn't have to do it one message at a time, writing the updated filter each
time. however, if the message cannot be parsed, it isn't trained during import,
and deliver would try to train it again. it would try to open the junk filter
to do so, but that would fail because the import function already has the
junkfilter open (and the timeout is reached). a missing error check would
continue with a nil junkfilter, resulting in the nil pointer deref.

this adds the missing error check, and makes sure the deliver function does not
also try to train unparseable imported messages.

report from Jens Hilligsøe
2023-02-26 22:21:13 +01:00
7e56ed9e8d Merge pull request #9 from belst/main
fix config options
2023-02-26 20:45:10 +01:00
8e178d9a1f fix config options 2023-02-26 15:57:28 +01:00
e07fb5c0f5 add roadmap to readme, and shuffle some sections 2023-02-25 13:38:07 +01:00
b80d85fd3f for smtp error responses, add the encrypted cid for easier debugging 2023-02-25 13:07:32 +01:00
b8fa918d74 add a bit more logging for non-SNI tls requests
for peace of mind.

these are probably requests to port 443 without SNI.
2023-02-25 12:56:25 +01:00
26fcaa17f5 consistently use lower-cased field names for logging lines 2023-02-25 12:37:59 +01:00
0ede7f78c1 add option to handle autoconfig and mta-sts requests without TLS, for when it is reverse proxied
for #5 with hints from belst & idnovic
2023-02-25 11:28:15 +01:00
ac3a3e496e make api urls relative, so they can work then being reverse proxied
for #5, found by belst
2023-02-25 11:20:07 +01:00
b1dcd73ebe help run mox with docker
in the Dockerfile, allow running on privileged ports and expose those ports.

add a docker-compose.yml with instructions for the quickstart.

fix running imaptest somewhat. after a short while it will hit the rate limiter.

in quickstart, recognize we are running under docker, and print slightly
different commands to set permissions, and skip generating the systemd service
file. als fix cleaning up the right paths during failure in quickstart.

for issue #3
2023-02-24 14:16:51 +01:00
210fd34702 umask logging line tweak 2023-02-22 23:26:00 +01:00
f2fd6241a0 do not bind to port 443 for tls-alpn01 if there is no ACME configured
for #2
2023-02-22 23:22:42 +01:00
54981fbc93 add a todo for preventing message content from being logged on common imap append to Sent 2023-02-22 10:43:21 +01:00
a480bcd583 use standard useradd options only, so the command works on bsds too, and no need for users to add an explicit -config flag anymore 2023-02-22 10:42:20 +01:00
8affa12c39 prevent panic when starting with a mox.conf that references an ACME provider that isn't configured
the error handling code attempts to collect all error messages it encounters,
to print them all at the end, so you can fix them all before trying again. so
we try to continue preparing the config after an error. in this case, we
continued as if acme.Manager was properly set and nil-dereferenced it.

for issue #2, from hismailbulut
2023-02-21 23:06:11 +01:00
cc3a2c9bc8 make it possible to serve https on a different port than 443 through configuration
so you can run mox on openbsd with port redirects in pf.conf.

in the future, starting as root, binding the sockets, and passing the bound
sockets to a new unprivileged process should be implemented, but this should
get openbsd users going.

from discussion with mteege
2023-02-18 16:53:06 +01:00
a289a3e771 when delivery fails due to missing 8bitmime/smtputf8 extensions, make it temporary failure
if you start delivering, openbsd's spamd will not announce 8bitmime support.
but once you get patched through to the actual mail server, it will likely
announce 8bitmime support.

perhaps we should also just attempt to deliver 8bit email without 8bitmime
extension. probably better for users than not even trying to send the message.
perhaps in a non-strict mode.
v0.0.1
2023-02-17 21:58:05 +01:00
4e5030ff38 fix output of "mox checkupdate", and specify changes to be from newest to oldest 2023-02-17 20:14:26 +01:00
cafccefad1 fix integration test, the delay for new senders was causing our db open to fail 2023-02-17 19:30:30 +01:00
6df4b454d5 update dependencies, including bolt with stability fixes 2023-02-17 18:55:01 +01:00
fb3794e31b only send \NonExistent for IMAP4rev2, and automatically subscribe to imported mailboxes 2023-02-17 18:35:11 +01:00
ad900b74e7 on mtasts error during delivery from queue, continue delivery but with strict tls
no need to not even try delivering in that case. if anyone is preventing mtasts
to work, nothing was achieved, because our fallback is the strictest behaviour
that could have been specified in the policy.
2023-02-17 17:19:41 +01:00
3608d0e246 more testdata to ignore 2023-02-17 17:06:27 +01:00
f76fe26976 when exporting mbox files, set new status,x-status,x-keywords headers, and remove any content-length header 2023-02-17 17:04:48 +01:00
c65731ac56 fix 3 cases of shadowed variables (other than "err")
2 of these were actual bugs.
2023-02-16 13:24:51 +01:00
5c33640aea consistently use log.Check for logging errors that "should not happen", don't influence application flow
sooner or later, someone will notice one of these messages, which will lead us
to a bug.
2023-02-16 13:22:00 +01:00
ef8e5fa1a8 on accounts page, fix opening destination with utf8
by decoding location.hash
2023-02-16 10:26:28 +01:00