Micro-outages do happen on IP.Networking says integrator

Telecom's ongoing problems with micro-outages have spilled over into its high-end DSL product suite, IP.Networking, despite Telecom's assurances that the problems are confined to JetStream only.

Telecom's ongoing problems with micro-outages have spilled over into its high-end DSL product suite, IP.Networking, despite Telecom's assurances that the problems are confined to JetStream only.

John Read, general manager of network services at systems integrator Sietel, says his company has worked on three network roll-outs in the past three months, all of which are experiencing micro-outages.

"It's like banging my head against a brick wall trying to get the message across to [Telecom] that this is going on."

Last week Telecom CTO Murray Milner told IDGNet that IP.Networking has "none of the issues" associated with micro-outages in the JetStream product family. Read says that simply isn't true.

" We're having exactly the same outages and exactly the same problem that you get with JetStream. It's all a load of rubbish."

Read says he has heard from some people within Telecom that DSL simply isn't up to the job.

"They're saying customers should change ADSL to frame relay and I know customers whose networks have been taken out and replaced with frame relay." Read says he's in a difficult spot caught between customers crying out for new technology and Telecom's silence on the problem.

"We've got some nasty issues out there and it's becoming a bit of a struggle."

Read places the blame squarely at the feet of Telecom's decision to use RIP or routing information protocol to maintain its routing tables.

"Essentially it sends the whole routing table around the network every 30 seconds and sends updates out. It's a basic protocol for routers to update each other with."

Read says because Telecom's network has so many different types of routers it loses some from the update cycle, leading to micro-outages.

"When they update themselves if they miss out on that routing table then you lose your route. It's not so much a fault, it's the type of protocol they're running is the wrong one."

Instead, Read says the protocol of choice would be OSPF (open shortest path first) which is more robust for a diverse environment.

"It sets up diverse paths through the network so you don't have things falling over."

In October, PC World columnist Juha Saarinen was told that Telecom would be switching over to OSPF protocol "in four weeks' time" and last week Milner confirmed Telecom is working on two separate fixes for the JetStream problem.

Milner was not immediately available for comment.

Related stories

RIP'ed routing turns JetStream into a notwork

No guarantees over JetStream fixes

Join the newsletter!

Error: Please check your email address.

Tags jetstream

Show Comments
[]