I’m new to * - we’re in the beginning of a global deployment over 6 offices - I was wondering how you guys manages and keep your passwords and voicemail configs managed.
I’m using pattern matching for my extensions, that helps somewhat but I’m sure at some point I’ll want to hand creating users to other people and I’d rather they didn’t edit the config files manually.
we have 8 servers, about 50 people per server, and do quite well just by hand editing the files (this is a company with very high turnover, so we’re changing 2-5 accounts per week per server).
we make extensive use of the #include directive, and have nost of our dialplan split out into several files, like so:
extensions.include - all non-static extensions mapped here
menus.include - all IVR menus mapped here
inbound.include - all inbound DIDs and scheduling defined here
outbound.include - all outbound dial patterns defined here
scripts.include - all macros defined here
it may not sound like it does much, but our extensions.conf is only about 50 lines or so total, and most of those are includes or static extension mappings, and the rest of the files are not much larger. it may not sound like it helps, but having 6 files with 100 lines each is much easier to manage than 1 file with 600 lines. it’s especially easier for including contexts in one another and whatnot.
that’s our method, and we can add a new user in under 30 seconds this way. we used to use thirdlane’s PBX manager, but it is actually faster to just hand-edit the files.
i used to be a big fan of AMP (with my mods), but now it just gets in the way. although i still use it for installations where they want to tinker themselves. with the recent revelation (to me anyway) about Realtime static i’ve started putting all my configs in a MySQL table and writing my own very basic interface. but then i want my system to run the way i want it to, and AMP/FreePBX doesn’t fit it.
Yeah, I think this will be the way forward in the long term. I’m ok with PERL and the company has plenty of .net devs so we should be able to sort something out.
I am a little surprised that there’s nothing fancy yet though. It looks like Asterisk is so close to fully beating many commercial switches but just a couple of point like this probably put a lot of potentials off. a great pity really.