Friday, May 12, 2006

RIPE vs route-flap damping

RIPE has issued ripe-378

It suggests that 2 major problems arise from implementation of route-flap damping:

1) Impeded convergence
Basically a ripple effect, caused by the flap penalty with a change in BGP attribute (in this case, a new path to the same AS). The prefix probably has not been withdrawn from the routing table at all, but route-flap damping can be triggered at the 2nd remove - i.e. at the neighbour of the router/AS making the routing change. Not mentioned, but tunable on some implementations, is that attribute changes can incur a lower penalty than straight prefix withdrawals. This could help alleviate local problems, but will start causing inconsistent (and difficult to diagnose) path selection.

2) Transit of updates
Related to the consistency of implementation of RFD: timing of routing updates. Because of differences in the Minimum Route Advertisement Interval (MRAI) Timer (the amount of time a router waits before passing on a route update), as well as differences in paths, multiple updates for the same 'flapping' AS can arrive at a router many hops away and cause multiple updates as the router evaluates the new routes available. This can easily trigger flap damping. There have been actual measurements where this resulted in a single prefix withdrawal producing 41 BGP events a few hops away!


Result: ripe-378 recommends against RFD, citing higher router CPU power as able to cope with flap events without damping.

Yay Stupid Network!

Wednesday, May 10, 2006

Big Brother's crackdown on crime

FM has an interesting sidebar in the current issue.

All SIM cards and phones that are sold or given away must have a paper trail: a record of the recipient's name and address and a photocopy (old skool!) of the recipient's ID!

Of course, post-paid (i.e. contract) users already have this in place courtesy of their Service Provider taking down their details to ensure payment.

"The measures are part of a security crackdown on the use of cellphones in crime , security officials say"

I wonder what the penalties for non-compliance are? Gareth, David, Herman: I'm talking to you here! You don't have photocopies of my ID and probably don't know my exact address. And I've got your phones! HAHAHAHAHA. I am a criminal genius!

PS Vodacom wants to use an electronic form for their customer registration. Mmmmm...electronic photocopies...

PPS [For extra credit] How many pre-paid cellphone users are there in South Africa?

Cell C number portability

Cell C has a small FAQ on Mobile Number Portability

Interesting points
1) MNP will take effect on 1st July 2006
2) No mention is made of offers to buy subscribers out of current contracts with competing networks. So much for customer acquisition. Maybe Virgin will be more proactive.

At least we don't have some of the crezzy US problems like GSM vs CDMA, BREW vs Java, etc etc

Thursday, May 04, 2006

LLU: PSTN vs VoIP

Analysts suggest that any company with leased telephone-number-space (i.e. currently to be Telkom, the SNO, MTN, Vodacom, Cell C, and the upcoming Virgin) will be able to connect to LLU customers and provide PSTN voice service.

Of course any (other) company can provide VoIP service. The crucial differences are:
1) Standard PSTN equipment (handsets & PABXs) vs VoIP equipment
2) ’Cheaper’ termination on PSTN calls (although e.g. IS can terminate on other networks, they don’t get anyone terminating on their network, so they don’t recoup anything as a percentage of calls) – reminds me of partial-settlement network peering
3) Inbound calls from the PSTN via standard dialling – basically the definition of PSTN/leased telephone-number-space

These deficiencies are not endemic to being a VoIP provider. VoIP refers just to the transport method, of course, but is traditionally used to imply an IP connection for the voice equipment too.

And the solutions to the above-listed problems aren’t too difficult (from a technological, if not policy, view):






1)Use Asterisk as a VoIP (service) to PSTN (handset) translating PABX. Comparatively low hardware cost depending on implementation size
2a)Technological solution: Avoid PSTN termination. As more users migrate to VoIP, Metcalfe’s Law comes into effect – more users means more value for users to migrate, which means even more users.
2b)Technological solution 2: Terminate on the cheapest destination interface (‘least-cost-routing v2’ – this pre-supposes ability to determine that multiple interfaces connect to one destination (either a sort of reverse-NAT, or a lookup, e.g. NAPTR) vs traditional hot-potato routing)
2c)Policy solution: Regulate excessive interconnect fees (that 600%!)
3)Acquire alternative numbering. ITU-T should demand that Enum compliance allows the various Carriers to dish out extra space (currently only a maximum of 10% of the number space is usable because of the way that area codes are assigned – 2 digits instead of 3) and that it is routed to by the PSTN. Policy questions will probably block this for a while, until ICASA decides whether to promote competition. Alternatively an international provider could supply globally-visible telephone number space and use NAPTR records to direct SIP sessions.

New Zealand to implement LLU

The Register has an article on New Zealand's desire to leave the bottom third of the log of broadband countries.

Gov.nz plans will do this by implementing LLU (local loop unbundling), increasing regulation, and promoting investment in new rival networks (including fibre and wireless).

Go Kiwis!