Slashdot Mirror
Ask Slashdot: Open Communications Set-Up For Small Office?
New submitter earthwormgaz writes "I've started at a small company and our phone system is crusty, old, and awful. We've got email hosted elsewhere on POP/IMAP, and we've got no groupware. The server here is Windows small business whatever-it-is and Exchange isn't set up, but I've put CentOS on it in a VM, and I'd like to do everything using open standards and open source where possible. I've been looking at SOGO, and these phones. What are my chances of getting all this stuff working together? What other suggestions have people got a for a small office and communications?"
if you're starting a business, just about the last thing you should be doing is worrying about is being sysadmin for your phone system - let alone doing so according to the "right" political principles and hoping you can get it to work together. Call your local phone company, get setup with them or some other turnkey provider and turn your attention towards your business.
If you don't mind paying for a product (and don't want to use Google Apps), take a look at Zimbra:
http://www.zimbra.com/products/index.html
It has an Outlook plugin so your Windows users will be happy, and it speaks Activesync, so any smartphone should be able to sync email contacts and calendar with it.
I haven't used Zimbra for a few years, but last time I used it it worked quite well -- much easier to set up and administer than Exchange, and cheaper too.
I've had good luck with PBX in a flash. You can run it on a small atom server for small numbers of people: http://pbxinaflash.net/
It works well with the Cisco SPA series phones: http://www.cedarpc.com/cgi-bin/commerce.cgi?preadd=action&key=24600
You can use things like SugarCRM and OpenFire with it. Share documents with MSOffice and a Subversion repository (you can probably even install SVN on the phone server). That's really all you should need to start a small company -- you don't have to think big yet, and when you do you should pay someone else to worry about it so that you can do the important stuff that goes with running a company.
Came here to say this. Use those phones with Asterisk, easy peasy. There's also Trixbox for the textophobic Windows admin.
"When information is power, privacy is freedom" - Jah-Wren Ryel
Asterisk may not solve all your problems, but if you are using VoIP phones and know Linux this might be an option. Plus it is open and fully customizable. Might be worth a look. http://www.asterisk.org/
If someone tries to kill you, you try and kill them right back
If you do it at no cost, the boss doesn't view that as a win for the company. He views it as suspicious, because where he comes from, spending is the key to getting somewhere, and everything costs something. In fact, he judges employees on how much they make (especially if he is new to the company), not how much they get done. It makes perfect sense to him, no matter what you think. To the boss, money is the bottom line to everything.
Let's put it this way. If your startup finds themselves hard on cash and needing someone to "temporarily answer phones", they will choose the person who makes the LEAST. (I speak from personal experience on this.) Why? Because his budget is tiny, and therefore "whatever it is he does" must not be as important as the person with the larger budget, or the person with the larger salary.
If a new IT guy comes along and spends twice as much as you, then the new IT guy is MORE valuable than you, not less. You will be considered the amateur, and he will be considered the professional, no matter how much actual "work" you get done.
So in conclusion, spend as much as you can, keep on spending as much as you can, and to hell with what actually happens to the company.
"I've started at a small company...."
1. Keep POP3/SMTP access; if necessary enable LDAP.
2. Use something like Google Apps for Business - includes e-mail (POP3/SMTP/LDAP) and Calendaring; $50/user/year.
3. Stay away from Outlook if you can help it; if you can't then at least stay as far away from Exchange as you possibly can. You'll save yourself a lot of headaches in the process. And if you can, enable your users to use Thunderbird (with Lightening if you want Calendaring); it can access LDAP and Directory Services for a unified address book too if you like.
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
...aaaand when Google changes its terms, or discontinues the product, you are well and truly hammered. Which Google definitely does do from time to time, and you can't predict when they'll decide they've had enough of supporting some chunk-o-freeware they cobbled up. Look at the wreckage they made out of Google base -- terrible, terrible support, and now they're converting to a "paid" model, which means that the product data you upload to them that they get to place ads all over... you now get to pay for. And there's plenty more like that.
Do NOT put your data "in the cloud." That's the very worst thing you can do. If you have a business, YOU should be in 100% control of your data and your backups.
The tech you use for documents should be chosen (1) so that you own the applications and (2) so that you can interchange any documents with others that you need to (color separations? Probably Photoshop. Writers and editors? Probably Word. Spreadsheets? Probably Excel. etc.)
You need a database? PostgreSQL or MySql (and I'd definitely go with the former... the latter has been, shall we say, "compromised.")
Just keep it to real applications that run under a real OS that you expect to be supported for some time. It will not be the least bit amusing to get that "end of life" notice from Microsoft or Apple or Ubuntu or whomever.
So many people here are assuming they understand your requirements better than you do, and those are the ones who could successfully parse TFS.
I run an opensource stack in-house because I need to customize what it does for my needs. None of the hosted products would work for me, and software freedom isn't something I throw under the bus for short-term gain. Currently it's a postfix/MailScanner/SpamAssassin/sqlgrey/dovecot/sasl/davical/asterisk/freepbx stack, but I've also never seen Sogo before, so thanks for linking that. I've been meaning to integrate Fumambol/SyncML and that does it built-in, so cool.
The other product I've considered is formerly-BBS-software Citadel, but I'm sufficiently suspicious of monolithic software to have not tried it out in production (the Unix way seems better). Sogo does more, though, so that raises the activation energy a bit.
On the phones side, I'm looking to replace the FreePBX system because it's increasingly buggy as new versions come out. There was a good interview with the 2600Hz folks on FLOSS Weekly recently about Kazoo. Their docs are very targeted towards a cloud-hosted version, which is fine, but I also haven't put in the energy yet to do a local install without docs. But it's on my very short-term list.
They seem to be headed in the right direction at least. Intergrating Sogo with Kazoo might be a nice direction and it doesn't seem like either community would be adverse.
Grandstream phones have the best bang for the buck, but aren't always quirk-free. That said, with a few tweaks they're very reliable and very cheap compared to Avaya. Their better models also embed linux, so I like to support them with my cash for doing so.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
The last time i worked in an office, there was no phone on my desk. If my boss wanted me, he IMed me.
You need to really be concerned about the following:
1.) Provisioning the equipment. I don't know how "small" a small office is, but this is going to spiral out of control quickly if you don't have an elegant way to setup handsets and make changes.
2.) Your change from circuit switched to packets. There are a lot of discussion points here, but the biggest you need to be aware of is latency is king. You might have a really slick p2p setup with OpenSWAN on 2 high bandwidth, cheap DSL or cable connections, but the jitter will kill you.
3.) How does your voice come in? If you are under contract and you have a PRI or some TDM circuit, you have to consider how you will interface that, and the cards you will need, or the SIP gateway you'll buy are not cheap.
4.) Who is going to manage the call routes, system secuity. I'm well versed with Asterisk, and you'll not find an all inclusive interface unless you go the Digium SwitchVOX route. If you don't pay close attention to security up front, you will experience toll fraud pronto.
5.) Handset support. What are you going to do for replacement parts, who is going to setup all the buttons, etc.
6.) Codecs. Some of the best are not free, i.e. G729. Just about any handset you get will support G711, but 12 bits of fidelity at 64k/sec each way (plur overheard for UDP/RTP) is not that great.
7.) Voice prompts, auto attendants, voicemail, etc.
8.) Status/BLF lights on phones. There isn't really a standardized way to do this, but SIP's Subscribe/Notify is used by some, I think Aastra.
9.) Key system habits. You won't be able to "pick up Line 2".
If I haven't scared you out of it yet, Aastra and Snom make excellent, RFC 3261 compliant handsets, Asterisk is a lot better than it used to be, and there are some alterntives you might find interesting like FreeSwitch or YXA.
Good luck.
mov ah, 4ch
int 21h
Just pay for hosted Exchange. Unless you're running an email company, you should not be worrying about what software your email/groupware is using. Save your high principles for when you're making a profit.
I don't respond to AC's.
My background is telecom and I have a lot of experience in that. My recommendation is to go with a hosted solution.
DO NOT INSTALL ASTERISK YOURSELF AND THINK YOU'RE GETTING A PHONE SYSTEM FOR FREE. You'll just waste time having to configure hardware, software, and dumb things like tuning analog POTS lines or wonking around with dial plans or something that you probably have no idea how to do.
Ok, back to the hosted idea. Let's compare the big costs with a traditional PBX and a hosted PBX:
1. Phones - you're really not going to avoid this cost. Budget $200 per phone set and be happy if you come in less. Remember, cheap phones are cheap for a reason. Spend the money and get a handset with a nice weight to it and a speakerphone that works well. If you get a traditional PBX like the Avaya system you looked at, there's a good chance you're looking at purchasing proprietary phones. If it's hosted, I recommend Polycom. Whether you have hosted or a traditional PBX, this will be one of your biggest costs.
2. The PBX itself will be a big cost. Avoid this by not buying one and going with a hosted solution.
3. The maintenance/service contract is the third huge cost, regardless of whether you go with hosted or traditional PBX. You're really not avoiding it with a hosted solution, in fact it might even be slightly more expensive, but you're paying for it month to month.
Since you can probably start small and grow into most hosted solutions, switch your conference phone over first and make everyone use it. You'll find out quickly if the call quality will work or if people have complaints.
Quality of service will be an issue with a hosted solution, so make sure you have bandwidth and if you need to set up real QoS on your router, know how to do that.
----- obSig
are radio frequency transmitting/receiving communication devices using digitized packet switching, which simulate copper-wire based telephone service but fails due to the lack of true full duplex and high latency.
For those of you who are too young to remember talking on a 20th century circuit-switching copper landline telephone system, I will describe the experience: it was like talking to another person in real life. You talk and they talk, sometimes simultaneously, and both parties could hear and understand everything... in real time.
I also remember gas was 95 cents a gallon back in 1995. Now get off my lawn.
Get everybody a cheap mobile phone with a business plan or agree to pay $50/month if they use their own phone (most people will). Move your email, calendar and documents to Google Apps or similar. And then focus on your real business.
If you've already purchased and using (albeit only barely) Microsoft SBS, take advantage of Exchange before you spend any more money on a new system, otherwise you're just wasting money. Exchange works quite well, obvious straight-forward connectivity with the Outlook client. Administering Exchange isn't the end of the world, and is actually quite easy in an SBS environment. I would suggest setting up an alternate internal smart-host (smart-relay) so that you don't have to expose the Exchange server directly to the internet. Courier MTA works VERY well (and is the exact setup we have internet->courier->exchange).
Setting up a Jabber IM server internally is easy as well, otherwise use Google Apps and have your email domain hosted there and just use Google Talk with the various AV plugins.
Setting up Switchvox (Asterisk) is a purchase, but I 2nd the comment by others to find you a local phone service retailer and let them deal with phone integration. If you do decide on a hosted solution for email and voice (voip) then make sure you don't skimp on the internet connectivity. I worked at a place previously convinced VOIP was the way to go, but management would cringe every time you talked about capacity of the external connection and the need to upgrade.
Just my 2cents...
think before you write, it'll save me moderator points.
Please make sure the system is well documented, easily maintained, professionally supported, and doesn't require a sysadmin's level of knowledge just to figure out how it works. Maintaining the phone system should only take up a trivial amount of my time.
Signed,
The guy they hired after you moved on
#DeleteChrome
I've not touched MySql in a long time...not since Oracle took it over at least.
Can you describe in what ways it has been "compromised" by Oracle?
Has no one forked MySql yet?
I have always preferred postgres for more hard core database stuff, but mysql always had its place....
Light travels faster than sound. This is why some people appear bright until you hear them speak.........
Google Apps for Domains is a paid product beyond a handful of users, and the services are Gmail, Calendars, Contacts, etc. Services that aren't going away.
I've spoken with their tech support, it's fine. Their import utility against exchange mailboxes works fine.
And with a huge percentage of small businesses still sending their email from @aol.com, I'm pretty sure the "zomg not the cloud" is falling on deaf ears. Small businesses want stuff that works. Google Apps works better than most local services, it's less expensive, has support, requires no maintenance, and isn't going anywhere.
So no, it's not "the worst mistake you can make." Here in the real world, we do what makes sense.
That kind of thinking is what leads to the very finest of vendor lock-in you could imagine down the road - and it's total bullshit. Investing a few hours of research and setup effort in a standards-based, transparent and reusable technical foundation for what is going to be the backbone of your company's communication both on the in- and outside for many years is definitely something to worry about - unless you have no problem whatsoever with buying your whole frickin' phone system all over again once you pick up the 11th employee, because the (cheap but proprietary) license and hardware you acquired when you started out "does not support more clients", or some such crap.
We just paid a few grand to extend our phone system from supporting <=50 clients to supporting 54 (and possibly more; even up to 70!!1!) clients. That's what you get from choosing the wrong solution in the first place, and if you let it become a vital component of your infrastructure - you'll have to stick with it and it will cost you dearly, because outright replacing it with a saner choice is always the more expensive one _in the short term_. Typically until the next forced upgrade cycle comes around.
:%s/Open Source/Free Software/g
YTARY!
First off Exchange is the most complicated and evil thing ms has ever made next to sharepoint. You dont need it! Here is why?
You dont just install it. The product actually alters AD itself at the schema level! So lets say you forget to raise the forest level in your domain as you just installed Server 2003. I bet you nooobs didnt know Server 2003 runs as Server 2000 forest and domain by default?! Somethin non win admins commonly make.
Oops just reinstall right? Nope AD has now been corrupted at the schema level and all users cant receive email anymore. Not even a tape backup can save you. Now imagine you have it working? How can people send you email? You get a ton of error messages when installing your cas outlook on the web about it not having a certificate?! Oh now you to create a Sans certificate. Now you need to register your web server so people can email you. What? You have to create a freaking IIS server too??
http://saveie6.com/
There's a fork of MySQL called MariaDB (http://mariadb.org/). I've been using MariaDB on my servers since Oracle purchased MySQL
> Has no one forked MySql yet?
Yes. The original author. Michael "Monty" Widenius. He named his first database after his oldest daughter My, the new one after the second daughter Maria.
http://mariadb.org/
Ignorance killed the cat. Curiosity was framed.
Cobbling together things rarely makes sense unless your time is free or you need something the various providers don't support.
IMO, unless you're going 100% open source for some philosophical reason, you can't beat the combination of Office365 and Windows InTune.
~$35 a month (O365 E3+InTune) per user gives you centralized desktop policy management, hosted e-mail, document sharing via Sharepoint, enterprise SA for Windows (so you can use/mandate Bitlocker, DirectAccess, and get free upgrades to Windows 8, etc), desktop software management (pushing out updates, new software, etc), Office Professional Plus, and Lync with telephony support. Another $20-$30 a month per user and you'll have direct dial in and out supported, with automated attendant, voicemail, and everything else, all in the cloud, all managed by one person via a web browser. Pay another $40 for your sales guys and you can flip on CRM. Hell, its worth it just to avoid dealing with all the "I forgot my password to our file sharing service" questions.
IMO, you could run IT comfortably for a knowledge-worker-centric small business with 30-40 people with one guy if you use the right infrastructure. And you won't have your infrastructure fall apart when the guy who cobbled together your stuff quits. A real small business and one person a couple hours a month could probably maintain it if they can follow directions.
Seriously, focus on your business, not this kind of crap.
This is the attitude I took away from the post as well. He immediately sees Windows whatever it is, and installs his preferred flavor of Linux and still doesn't express any idea of what he's supposed to do other than something with the phones and maybe email something or other. If Windows is there, someone's already paid for it. Use it unless they don't want it or it's a woefully old version.
I for one do not believe he should touch the phone system, that one is best left to a specialist company or package. As soon as there are issues with them he's going to have everyone in that office up his ass to get them fixed and to make sure they work right all the time.
And for god's sake, lose the attitude. You're generally going to get responses from people here with the same attitude, but don't take it for granted because no one else outside the IT realm gives a flying fuck about your disdain for windows.
Comment removed based on user account deletion
...doesn't mean you should.
I work for a company that provides VoIP and data services (the kind of company the OP should be calling). We have some damn good sysadmins, who can run everything from Asterisk to Postfix. But we don't. Internally, we use an off-the-shelf Asterisk implementation with a nice interface, and for email, hosted Exchange on one domain and Google stuff on the other. We could write our own ticketing system, but we use a hosted solution. It costs us far less to pay for some of these hosted services than to develop our own.
So, while we could hand-roll everything, we don't. Our whole business is based on Asterisk somehow, but we don't use it raw. Time spent maintaining our own software is time we're denying our customers.
As far as boxed, premise solutions go, I really like the Adtran 7100. Handles all voice and data up to 100 seats, and their phones are very good too.
That I'm right, and you don't like it, doesn't mean I'm a troll.
The problem with the "keep your data and apps where you can see them" approach is that the TCO is horrendous.
Yeah, Google's cloud applications suck. That's due to Google ADHD issues, not the fact that it's cloud hosted. Tell me you've never been screwed over by a traditional application whose publisher lost interest in it.
It's perfectly true that some cloud applications are too immature and not ready to replace their traditional counterparts. Office applications (word processors, spreadsheets, etc.) are certainly there, at least for serious users. But the best CRM and HR solutions are cloud-based, and have been for some time. And the companies behind them are here to stay.
ok, so you've just turned on virtualization on your SBS server. you just broke it. Microsoft supports SBS installed as a GUEST but not as a HOST for virtualization. this is all over the microsoft knowledge base and the SBS Blog (blogs.technet.com/b/sbs) you'd best read up on SBS Best Practices before you make your server any worse. www.sbsbuilddoc.com
I'm not convinced that the TCO is "horrendous" and I think that line is used too much to try to sucker people into using a subscription/cloud base service.
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
Yeah, go with what's proven; whether that's hosted or local.
Agreed. But if you have a choice between a proven local app and a proven cloud app, you go with the cloud app to save money. That's a decision that the CEO of HP just made by ditching existing CRM and HR applications for their cloud equivalents. Which is ironic, because she's assuming (correctly) that using the very PCs and servers she sells for this purpose is too expensive
I am guilty of posting after a few beers so mod me down as I deserve ...
Slashdot is definitely the wrong place to pose this question as you may have already realised - if an open source god didn't exist then slashdot would have invented one
My two cents? Take the path of least resistance. Choose the best thing you think you can successfully deliver on time to make the business work. When you have more time you can make things work better - if they don't work in the first place then no one will care about your well intentioned ideas
At the same time - don't give up on your aspirations, just remember, deliver something you can control in the time given - you probably already know what that is.
Actually I have friends who own and operate small businesses. Their professions include court reporting, lawyer, restauranteur, Lloyd's surveyor, independent insurance agent, landscaper, painter, general contractor, dentist and marine transportation firm. Some are small less than 10 employee operations, but others are in the 50+ employee size. They have no problems with the equipment they have and the software packages that they have purchased. You do realize that there is a cottage industry that create vertical market software for these professions? They've been generally satisfied with their service and they like the fact that they control where and how their private information is handled.
Most of those businesses would still need pretty much the same amount of computer equipment and the only thing the "cloud" promises them is always working and up-to-date applications with no need to backup. This pretty much taken care of by the maintenance contracts that they have with their software provider anyway. I always stick to the current version of the software until it has been end-of-lifed. No sense worrying about file incompatibilities or having to update the workstations prematurely just for a few fancy and unrequested features. The backups are easily handled by the owner or one of their employees. I recommended that they keep a set of backups off-site at home or better still a safe deposit box at a bank.
This isn't rocket science, and we don't need to employ computer scientists to do the mundane stuff just because a computer is involved.
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
Having managed both Cisco Unified Communications and Elastix, I would say the babysitting is about equal, and the ability to respond to change is hands down in the Linux side. Cisco works well is everything is Cisco, and you are under a support contract. Get some version mismatches, and old VPN clients, and the Mac in the mix, and you better get a bottle of Tequila as well! What, you want to upgrade the Exchange server? Well, you will have to upgrade all the Cisco software for that...
I'd recommend running an XMPP server to provide instant messaging and more on the local network. I recommend Openfire for the server, it's fairly easy to get up and running and is Apache licensed; the server runs on Linux or Windows. It supports LDAP for authentication against an Active Directory network for user accounts so it will integrate well with your existing Windows domain. Functionality depends a lot on the client you select, but I'd recommend Jitsi (formerly SIP Communicator) which is very similar in many respects with Microsoft Lync; it is LGPL and supports enterprise features like voice/video calling, SIP integration, automatic provisioning via URL, encrypted connections, and a lot of other interesting features. It runs on pretty much anything, If you add Asterisk to the mix, you can tie Jitsi into that as well and get phone system integration and dialing from your desktop.
This does not solve the problem of Sharepoint, however.