Whole murder (2.3.x) setup on a single machine: is it possible?
ph rhole oper
slitbit at fastmail.fm
Fri Apr 21 05:39:33 EDT 2006
I need to make a whole murder setup on a single machine.Is this
possible?
According to the documentation, you can run the master mupdate process
on a frontend: "One machine to become the MUPDATE master server.
This can be the same as one of your frontend servers.Note that you can
have the MUPDATE master be one of your frontend machines, just do not
configure a slave mupdate process on this machine.."
In addition, "any one "unified" server can both proxy and serve local
mailboxes (proxy functionality in proxyd and lmtpproxyd has been merged
with imapd and lmtpd respectively"..wich leads to the logical
conclusion that you can have a single machine, running
the mupdate master process only, and set mupdate_config: unified, so
that it will act both as a proxyd and an imapd backend, as well as
a lmtpproxyd.
But while i was taking a look at the source, i found this piece of code:
mupdate.c:
if (masterp &&
config_mupdate_config == IMAP_ENUM_MUPDATE_CONFIG_UNIFIED) {
/* XXX We currently prohibit this because mailboxes created
* on the master will cause local mailbox entries to be
propagated
* to the slave. We can probably fix this by prepending
* config_servername onto the entries before updating the
slaves.
*/
fatal("can not run mupdate master on a unified server",
EC_USAGE);
}
If im guessing right, you cannot.But what if im not using a slave
mupdate process?Should i comment out the above piece of code and expect
it to work?
Additionally, for those who are wondering what use would a
single-machine murder configuration be, im trying to integrate Cosign
(Web SSO) with the HORDE framework.Because IMP (the webmail client) does
not support GSSAPI authentication to imap, im planning to setup a local
imapd proxyd using the alleged -N option (wich by the way i don't see
nowhere!), wich means it accepts passwordless connections.
simon.
--
http://www.fastmail.fm - Or how I learned to stop worrying and
love email again
More information about the Info-cyrus
mailing list