Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Ascend Communications (ASND) -- Ignore unavailable to you. Want to Upgrade?

To: bucky89 who wrote (45416)	4/27/1998 5:46:00 PM
From: The Phoenix	Read Replies (1) \| Respond to of 61433

OK Bucky... I've got a minute here... Once again this has been an interesting conversation and the funny thing is is I think we're in violent agreement. I know Cisco looks to IP over Sonet as the "holy grail" but reality tells us that ATM will play a significant role in the SP networks on the future. In the end I think we both agree that IP edges and ATM core's are where service providers are taking their deployment so perhaps this conversation is mute. Nonetheless it has been interesting and hopefully educational no only for us but for those that have taken the time to read. With that I guess it's my turn to respond. Gary takes Bucky's wicked smash and slaps it back with a backhand deep to the baseline. There are two weaknesses I'm thinking of with a connectionless protocol. The first you already guessed--the need for every router to examine each and every packet or flow... Yes, but layer 3 switching eliminates this. In fact, Cisco has been doing cut through switching in their routers for a while now. The second weakness is a connectionless protocol cannot do absolute QoS on an end-to-end basis. Latency and performance from source to destination can only be measured and guaranteed in a relative sense. If a network becomes congested, then all traffic will suffer, although to varying degrees. For low priority traffic, performance will be terrible or nonexistent. For high priority traffic, performance will merely be bad. RSVP cannot do something like guarantee 50ms of latency for a particular type or path of traffic regardless of the congestion in the network. ATM can do this very easily. Well, I think you may be wrong about IP's ability to prioritize and guarantee QoS. There's been a lot of work in this area. But even if your not, I'm interested in why you believe ATM can "guarantee" specific end to end latency. I agree that ATM can guarantee jitter due to cells...but surely you're not saying ATM can guarantee delay under network outages or congestion scenarios.... Yes, you can "tune" buffer depths, and "throw away" packets, but retransmission will excerbate the problem. This is where best effort is superior. This DEN you're refering to is something new that I haven't heard of. Sounds like an extension of RSVP. Can't competently comment on it, but on the surface it doesn't sound like a scalable technology either. You say it's built on layer 7, but if a network is congested how is the router going to communicate with the server to make the prioritizing decisions? DEN is nothing like RSVP - it's a server based mechanism where, upon sign on to a network the server communicates with the network devices, assigns priorities, access, and security - all based on the application the user is running. That is, a user may have high priority for a video conference and a lower priority for email. When he sets up his video conference the server commincates with the network, sets up the best path, and assures priority/latency. This off-loads the routers from having to do this work. If there is a network outage the routers still have the capability to re-route the traffic... does the router have to query the server for every one of these flows? Or if the router caches the decision data, then how much more memory needs to go into the router? Router doesn't query, it's interrupt driven. No more memory required...in fact one could argue less. The possible routes are no different than if the server didn't exist, however now the router doesn't have to worry about communicating with various network devices...only one - the local DEN server. Note that there can be multiple DEN servers in a network for resiliency. You ask why I say RSVP will not scale in ISP networks. I say just ask the ISP's. Interesting considering they haven't yet implemented a full RSVP based network yet. ISP's aren't making money and can not afford to deploy perhaps. In the meantime other vendors such at QWST, USWest, and AT&T will indeed offer IP based nets...probably with ATM cores... It's not about the space it takes up in the IP header, but the amount of signaling that must take place among the routers and the number of states that must be maintained in a large network. Again, another interesting opinion I haven't heard. States maintained??? Routers have been managing many more "states" than ATM for years. Now however the transition of states (which has been troublesome) is being offloaded to DEN for management of the changes. The router becomes the brute force mechanism. As for IPS implementing RSVP or some form of layer 3 QoS... perhaps this will change soon. I can't say anything more than that right now.... Not sure if you're familiar with the MPLS strategy, but it is to read the IP header once and only once at the ingress to the ATM cloud, then switch the traffic to the egress of the cloud through use of labels and a pre-existing VC. Within the cloud, ATM will be able to prioritize better than IP because it is a connection-oriented technology, which we discussed earlier. Also, there is a HUGE improvement in latency and packet loss, versus IP routing. Yes, Bucky, I'm very familiar with MPLS. And yes...just like any protocol connecting to an ATM fabric the header is read only once and passed end to end through a managment intensive connection oriented protocol architecture...over rigidly constructed VC's. Not very flexible. Nonetheless the prioritization of this environment is only as good as the information in the IP header. Furthermore ATM is not able to offer multiple prioritization for multiple types of IP traffic. Finally I beg to differ with you comment about packet loss. ATM does NOTHING to protect against packet loss. In fact it relies wholly on the IP device to manage this process which is one of the ways ATM and FR reduce overhead... Finally, ASND knocking off CSCO as a leading networker is an interesting point of view. I'm not sure there are many that would hold such a view in the industry but I guess anything is possible. Problem is ASND is a narrow thinking company with a management team that just not that market savvy....well at least in comparison to Chambers, Listwin, and the gang at Cisco. In addition Cisco has shown incredible resilience and an ability to move to the issue of the time. Finally ASND is in a market window where they should be excelling...cisco's products are long in the tooth and they haven't had any major announcement in a while. This leads me to believe that something must be on the near term horizon. I will say it appears ASND has some recent wins that would lead one to believe that an "overthrow" is possible..but in reality I think this is a pipe dream and that Cisco will react...agian JMO.... BTW: ASND is a great company and I think this will be a good year for them. I do think they will need to partner soon however in order to continue to grow and become a 'tier 1' company. Afterthought...just saw this.... Here' they come...although this is campus based I have a feeling that there's more coming.. biz.yahoo.com OG