[VOIPSA Best Practices] Best Practices document structure set -next question: are these the appropriate areas?

Fri Jan 19 15:35:41 CST 2007

Hi Dan,

Ad fail-over: I think it would fit to section #2, Call Control. I think we could describe in that section the fail-over backup solutions. For SIP based systems we could talk about the efficiency of multiple DNS SRV records (HA vs. HP), and about sharing the registration information from UAs between multiple registration servers. So if a UA registers to a SIP server with higher priority which after some time broke down, then other servers with lower priority will be used for call processing. But then it is important to share the registration and call state information between those servers. 

BTW: It looks like the Call Control section will be a quite huge section :)

Ad QoS: Designing a secure voip network means the L1-L7 security but it also means a well designed a network with QoS. I think the document should show the best practices in securing the voip on L1-L7, but it should also show some of design paths on QoS – RSVP, CoS, ToS, DSCP, ....

Regards,

   Jozef

From: dan_york at Mitel.com [mailto:dan_york at Mitel.com] 
Sent: Friday, January 19, 2007 8:18 PM
To: Nhut Nguyen
Cc: bestpractices at voipsa.org; Jozef Janitor
Subject: RE: [VOIPSA Best Practices] Best Practices document structure set -next question: are these the appropriate areas?

Nhut, 

(Sigh) Yes, you're right.  We probably need a Section 9 called "Emerging Areas of Concern" or something like that[1] into which we lump SPIT and other miscellaneous things that don't yet have solid best practices (because they're not yet solid threats, IMHO) if only so that we can avoid the inevitable questions "Why is it that in your entire Best Practices document you don't mention SPIT?  How are we to guard against it?"   

Jozef, 

- I don't know where fail-over fits. Section 8, in *my* mind (but I'm just one contributor) was about all the physical layer stuff that many of us network geeks generally try to ignore such as backup power, physical security, etc.  I was thinking of it from the "availability" point-of-view in that if you lose power and don't have your IP-PBX *and* network switches powered by a UPS, your phones are nice pretty bricks.  And you should have them in locked rooms, etc.  I don't know that we'll have all that many Best Practices in this area. 

As to fail-over, I'm not sure... is it in Section 8?  Or is it in call control?  I'm not sure on that one... any other opinions? 

- Yes, caller ID spoofing is something to be addressed.  Probably in Call Control? (Eric Chen, now's probably a good time to weigh in before we assign everything to you! :-)    As for ENUM, there's probably a large debate on that one, but as we write the best practice around caller identification the debate can be waged then. 

- QoS - Yes, there is definitely a balancing act between quality and security... although I'm not entirely sure how we create that as a best practice.  Would the best practice be something like "Install and use quality monitoring equipment" - somewhat along the lines of what Gary was suggesting earlier? 

Love all these comments... keep on sending them in... this is *your* document as much as it is mine or anyone elses.  Let's make sure it's right. 

Thanks,
Dan 

[1] Or the section can be called "Issues Related To VoIP Security That Are Over-Hyped By The Media Because They Get Attention" (and yes, I know there are probably several members of the media subscribed to this list)

"Nhut Nguyen" <nnguyen at sta.samsung.com> 

01/19/2007 12:35 PM 

        To:        "Jozef Janitor" <jozjan at cnl.tuke.sk>, <dan_york at Mitel.com>, <bestpractices at voipsa.org> 
        cc:         
        Subject:        RE: [VOIPSA Best Practices] Best Practices document structure set -next question: are these the appropriate areas?

Hi All, 

Great discussion! 

Regarding SPIT I am not sure if the industry has any good practices yet. I believe this problem is emerging but solutions are still in the research stage. However, it may be a good idea to have a stub or a place holder to document emerging practices like these, or those that address security issues of presence information, for example! 

Cheers, 

Nhut 

  _____  

From: bestpractices-bounces at voipsa.org [mailto:bestpractices-bounces at voipsa.org] On Behalf Of Jozef Janitor
Sent: Friday, January 19, 2007 11:23 AM
To: dan_york at Mitel.com; bestpractices at voipsa.org
Subject: Re: [VOIPSA Best Practices] Best Practices document structure set -next question: are these the appropriate areas? 

Does the point #8 describe also the fail-over procedures? So if my primary call controller broke down then the second call controller will automatically handle the active calls? This may involve some clustering techniques. 

Also important part of voip security is the credibility of an incoming calling number (callid). Because in the PSTN network normally I can’t change my callid, but in the VoIP it’s usually not a big problem. Maybe this problem could be handled with ENUM. 

QoS – this is very important when we are going to talk about security in voip. We have to consider what level of security do we need because the encryption of traffic and other security features are always adding additional delay to the transported voice. If the value of this additional delay is going to be too high then our call is maybe for 100% secured but it’s not pleasant to hear. 

And SpIT. We know that controlling SPAM in our emails is very difficult . But controlling SPAM in VoIP will be even more difficult. So I hope that some pages in this document will also cover the possibilities of solving the SpIT problem. 

All the best, 
   Jozef Janitor 
    <http://www.cnl.tuke.sk/> www.cnl.tuke.sk 

From: bestpractices-bounces at voipsa.org [mailto:bestpractices-bounces at voipsa.org] On Behalf Of dan_york at Mitel.com
Sent: Friday, January 19, 2007 11:58 AM
To: bestpractices at voipsa.org
Subject: [VOIPSA Best Practices] Best Practices document structure set - next question: are these the appropriate areas? 

Best Practices team, 

Thank you to those of you who sent in comments either on the list or directly to me.  A special thanks to Eugene Nechamkin who took the time to write up a counter-proposal. Outside of his contribution, basically all the feedback was for proposal #2, structuring the document around functional areas, and so I'm going to say we're going with that. 

Now, the next question - is this list below from the wiki the appropriate list of areas for VoIP-related best practices? 

1.        Securing Voice and Media stream 
2.        Securing Call Control 
3.        Securing Management Interfaces and APIs 
4.        Securing PSTN Interfaces and Traditional Telephony Issues (i.e. don't forget toll fraud) 
5.        Securing Servers and Operating Systems 
6.        Securing IP Endpoints (ex. sets, softphones, etc.) 
7.        Securing the TCP/IP network (ex. VLANs, 802.1X, wireless, etc.) 
8.        Physical Security, including backups, power, etc. 

Are we missing any major areas?  Should these be modified or tweaked? 

It seems to me to be a complete list, but then again, I wrote it, so of course it would.  Any feedback is welcome. 

Regards,
Dan 

-- 
Dan York, CISSP
Dir of IP Technology, Office of the CTO
Mitel Corp.     http://www.mitel.com
dan_york at mitel.com +1-613-592-2122
PGP key (F7E3C3B4) available for 
secure communication 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://voipsa.org/pipermail/bestpractices_voipsa.org/attachments/20070119/58b3018b/attachment.htm>