6f51b49671
remove the hard-coded migration ID from setup.sh
2014-07-10 12:49:19 +00:00
41b3df6d78
manage hostmaster@ and postmaster@ automatically, create administrator@ during setup instead
...
closes #94
2014-07-09 19:30:17 +00:00
22a010ecb9
say that certificates are valid too in output
2014-07-09 16:38:56 +00:00
659b5c8aa3
if the server certificate can be used for a non-primary domain, use it
2014-07-09 16:38:42 +00:00
6c70b10c15
tell users to restart nginx after plugging in a new cert
2014-07-09 14:05:59 +00:00
deebda06e1
utils.sort_domains wasn't right
2014-07-09 12:35:12 +00:00
1a74b81f44
new nginx configuration yaml file to allow proxying of whole domains elsewhere
2014-07-09 12:31:32 +00:00
04e30ffa78
check that the installed certificate corresponds to the private key
2014-07-08 15:47:54 +00:00
10fbb2b293
in cf7053c124 I allowed editconf.py to insert a setting where we find it already commented-out in order to get an nginx configuration line in the right place, but it wasn't quite right because when run again we would insert the setting a second time
2014-07-08 00:48:22 +00:00
3bab63d4ce
update to Roundcube 1.0.1
2014-07-08 00:37:53 +00:00
7339bf080a
add a web_update script to trigger writing nginx config
2014-07-08 00:34:38 +00:00
59a9d02fa5
check that installed certificates are for the domains we are using the certificates for
2014-07-07 12:06:11 +00:00
3d4eadd436
the new migration management in c8856f107d left out the part where we actually keep the system's current MIGRATIONID... it was being lost when setup/start.sh was re-run
2014-07-07 11:29:21 +00:00
cf7053c124
set nginx server_names_hash_bucket_size to 64, fixes #93
2014-07-07 11:23:41 +00:00
430b2dec11
update default www page to link to the website, fixes #96
2014-07-07 07:07:54 -04:00
ad3f6f8424
adding externals and .env to gitignore
2014-07-07 07:06:36 -04:00
65fb65ada7
an mx record may be missing if the A record matches the A record of PRIMARY_HOSTNAME
2014-07-07 02:35:45 +00:00
28e254fb84
whats_next: Allow the PRIMARY_HOSTNAME to not have an MX because the default value means the domain itself, which is what we want anyway
2014-07-07 02:35:45 +00:00
e898cd5d2a
whats_next: wrap output to the actual width of the terminal
2014-07-07 02:35:45 +00:00
6a231d4409
clarify that an SSL cert can remain self-signed on the non-primary domains if the domain isn't being used for web
2014-07-07 02:35:45 +00:00
dcce98f84b
and remove the old documentation now that there is documentation on the website
2014-07-06 11:57:57 -04:00
05664f0a3b
have the README refer to the website for details
2014-07-06 11:31:17 -04:00
49d5561933
when adding/removing mail addresses also update nginx's config
2014-07-06 12:16:50 +00:00
c8856f107d
migrate the SSL certificates path for non-primary certs to a new layout using a new migration script
2014-06-30 20:41:29 +00:00
06ba25151f
get_domain_ssl_files returned the wrong path for the CSR for PRIMARY_HOSTNAME
2014-06-30 19:49:41 +00:00
b5aa1b0f31
walk the user through choosing the PRIMARY_HOSTNAME by first asking for their email address
2014-06-30 10:20:58 -04:00
fed5959288
s/PUBLIC_HOSTNAME/PRIMARY_HOSTNAME/ throughout
2014-06-30 09:15:36 -04:00
573faa2bf5
install the backup script as a daily cron job
2014-06-26 10:46:22 +00:00
87f001a5d5
some comments
2014-06-24 03:24:41 +00:00
f8cd2bb805
typo: www/default/index.html would be overwritten if it already exists
2014-06-23 19:43:19 +00:00
1dec8c65ce
move the SSH password login check into whats_next.py (it used to be in start.sh and then moved to an unused script when it became a problem for Vagrant)
2014-06-23 19:39:20 +00:00
d4ce50de86
new tool to purchase and install a SSL certificate using Gandi.net's API
2014-06-23 10:53:29 +00:00
30c416ff6e
rename the new checklist script to whats_next.py
2014-06-23 00:11:24 +00:00
5aa09c3f9b
let the user override some DNS records in a different way
...
Moved the configuration to a single YAML file, rather than one per domain, to be clearer.
re-does 33f06f29c1
2014-06-22 19:33:30 +00:00
45e93f7dcc
strengthen the cyphers and protocols allowed by Dovecot and Postfix submission
2014-06-22 19:03:11 +00:00
343886d818
add mail alias checks and other cleanup
2014-06-22 16:28:55 +00:00
deab8974ec
if we handle mail for both a domain and any subdomain, only create a zone for the domain and put the subdomain's DNS records in the main domain's zone file
2014-06-22 16:24:15 +00:00
4668367420
first pass at a management tool for checking what the user must do to finish his configuration: set NS records, DS records, sign his certificates, etc.
2014-06-22 15:54:22 +00:00
ec6c7d84c1
dont ask for a CSR country code on second runs because the CSR is already generated and any new country code won't be used anyway
2014-06-22 15:36:14 +00:00
8076ce4ab9
Merge pull request #74 from mkropat/mgmt-auth
...
Add authentication to mailinabox-daemon; resolves #67
2014-06-22 11:36:04 -04:00
9e63ec62fb
Cleanup: remove env dependency
2014-06-22 08:55:19 -04:00
d100a790a0
Remove API_KEY_FILE setting
2014-06-22 08:45:29 -04:00
554a28479f
Merge remote-tracking branch 'upstream/master' into mgmt-auth
...
Conflicts:
management/daemon.py
2014-06-21 21:29:25 -04:00
064d75e261
Merge pull request #73 from mkropat/syslog-logging
...
Tell Flask to log to syslog
2014-06-21 21:22:27 -04:00
e70bc50432
README parallel sentence structure
2014-06-22 00:34:49 +00:00
bb394242ef
Update documentation to use API auth
...
The updated instruction is not very user-friendly. I think the right
solution is to wrap the `/dns` commands in a `tools/dns.py` style
script, along the lines of `tools/mail.py`.
2014-06-22 00:07:14 +00:00
88e496eba4
Update setup scripts to auth against the API
2014-06-22 00:02:52 +00:00
447399e8cd
Update mail tool to pass api key auth
2014-06-21 23:49:09 +00:00
067052d4ea
Add key-based authentication to management service
...
Intended to be the simplest auth possible: every time the service
starts, a random key is written to `/var/lib/mailinabox/api.key`. In
order to authenticate to the service, the client must pass the contents
of `api.key` in an HTTP basic auth header. In this way, users who do not
have read access to that file are not able to communicate with the
service.
2014-06-21 23:42:48 +00:00
53e15eae15
Tell Flask to log to syslog
...
- Writes Flask warnings and errors to `/var/log/syslog`
- Helps to debug issues when running in production
2014-06-21 23:25:35 +00:00
Joshua Tauberer
Michael Kropat