I think you'll start to see this in Large Enterprises over the next few years. With the costs of supporting PC's so high, it would be easier and cheaper to use thin clients/thin wireless tablets that access a grid of virtualized compute resources and a SAN to run their apps.
...If its just an intern program for the younger generation, that great! But from an output perspective, what have we received from it? Is it worth the hype?
There are plenty of online exchanges for info and ideas, even problems that need solutions that link corporate R&D, Universities, and individual scientists. This is how a lot of companies innovate today - through collaboration, called C&D.
Modern humans's alone changed a lot over the span of 100,000's of thousands of years and are NOT stable over the span of 2 million years. Same with animals. Using the Galapagos islands as an example, they are estimated to be anywhere from 700,000 to 3.5 million years old. During that time, we can guess that they didn't always have animals on the island, so animals on that islands are considerably less than 2 million years old. During that time, the animals HAVE evlolved and changed from their mainline species. Sequences of animal records on that island show the change in a much shorter time span.
Wrong? You pointed out a 30 year old theory which never caught on with true scientists, yet caught the eyes of creationists because it could show a pattern of species appearing, as if by magic or some intelligent being. Even this article mentions all the controversy surrounding this crackpot theory. Go troll somewhere else.
..that happens over time. We just happen to dig up random fossils and see dramatic changes from the previous, older species. We forget that there were sometimes 10,000's or 100,000's of years in between the two species.
There isn't one "link" between two species. A situation where one day a parent gives birth to a dramatically different, more advanced offspring that is more evolved then the parents doesn't happen. And even if they was a missing link, the chances of that fossil surviving and us finding itwould be near impossible.
I'm not so sure we have too much money in VC's. I agree that we did in the 1990's, but there are far fewer ventures out there today. Most money is socked away in hedge funds these days.
Or they've already maxed out the market share in their current area's of competition. They have 80% market share in Routing and Switching, 60% in wireless, 65% in Linksys, they are now #1 in enterprise voice, #1 in director-class storage, etc... Chambers once said that if they tried to get to 100% market share in Routing and Switching and took all their competitors market cap value, Cisco's market cap price would rise a paltry $25b, nothing investors are going to jump at. Plus it would be a dog fight for little profit or $$. So they need new areas to grow.
Bottom line is that Cisco consolidates networks (consolidated the hodge-podge of networks in the 90's - IPX, DECNet, Banyan Vines, SNA, etc) onto IP and is looking to doing the same today (SAN's, Voice Networks, etc). Analog video is just another network out there that can be put onto IP (imagine all the gear in just one high rise in NYC) and you can see where Cisco's going.
Those all sound like Layer 8 issues (people, support and configs), not HW issues:) Granted HW and software fail on any box from any company, but in my experience most issues are caused by the operators, not the equipment. You can always outsource support of your network if L8 becomes too much of an issue - there are plenty of people out there who know that gear and can do it at a cost effective price; it would be hard to find people to support open source boxes at a cost effective solution...
Can one linux box become a high performance router plus (summary of just new security features released this quarter, never mind all the functionality released the past 3 years)-
Stateful FW Failover
Zone-based Policy Configuration
Cisco Unified Firewall MIB
SSL VPN, including support for Cisco Secure Desktop (Zone-based Policy Configuration means that administrators will be able to group physical and virtual interfaces into security zones to allow for simplified configuration of firewall rules. Firewall policies can then be applied to a zone rather than an interface. This will also simplify the process of adding or deleting interfaces on a router).
This is just a list of the NEW features released THIS QUARTER:
* ACL Manageability
* ADSL HWICs
Introduces 2-port ADSL HWIC Hardware.
* ANI Suppression During L2TP Setup
* Certificate - Complete Chain Validation
* Cisco IOS Firewall MIB
* Cisco IOS IPv6 Configuration Library
* Cisco Modem Relay
* Cisco Text Relay for Baudot Text Phones
* Control Plane Logging
* DHCP Option 82 per Interface Support
* DHCP Relay Accounting
* Dynamic Frequency Selection and IEEE 802.11h Transmit Power Control
* Easy VPN Server
* Fax Relay Support for SG3 Fax Machines at G3 Speeds
* FHRP - HSRP Multiple Group Optimization
* Flexible Packet Matching XML Configuration
* In-Service Updates to Gatekeeper Zone Prefix Configuration
* Interface Input Queue Unwedging
* IOS Firewall Stateful Failover
* IP SLAs ICMP Jitter Operation
* IP SLAs--LSP Health Monitor
* IP SLAs RTP-Based VoIP Operation
* Management Plane Protection
* MGCP NAS Package LAPB-TA
* MPLS Embedded Management--LSP Ping for LDP
* MSCHAP Version 2
* NAT ARP Ping
* NAT SCCP Fragmentation Support
* Network Admission Control: Agentless Host Support
* New Voice and Telephony Features in Cisco IOS Releases 12.4T
* OCSP - Server Certification from Alternate Hierarchy
* OER Voice Traffic Optimization
* OSPF Enhanced Traffic Statistics for OSPFv2 and OSPFv3
* OSPF RFC 3623 Graceful Restart Helper Mode
* OSPF: SNMP ifIndex Value for Interface ID in OSPFv2 and OSPFv3 Data Fields
* Packet Mode Services on D Channel
* RIPv2 Monitoring with SNMP Using the RFC 1724 MIB Extensions
* RSVP Agent
* RSVP Application ID Support
* SCCP PLAR with DTMF Ou
The game has long since moved from just forwarding packets to providing intelligence in the network. Now companies want integrated security, voince, application intelligence and application (l5-L7) optimization, QOS, high availability, etc.. none of which you'll find in an open source router. This is why the networking companies stay in business. If companies wanted cheap packet forwarders, they would have bought linksys, 3com, huawiei, hp or any other me-too commodity router. They didn't and Cisco won.
This type of model can be inherently secure as most of the processing, policies and intelligence for these types of applications will be in a central data center. The main difference between this new world and the old is the ability to access the compute resources from anywhere or virtually any device on any network. So if we use a lot of the lessons learned from the "olden days," we can create the best of both worlds.
They have come up with some pretty unsound.com business idea's lately - parallel internet, free wifi for the whole world, etc... I like Google, but it's getting silly.
Nothing new. From the propaganda side, we've been doing this type of stuff for years - Voice of America, for radio, etc... This is just a new medium. From the disruption of service side, we've also been doing this for years, most recently we debated weather to knock down Al Jazeera.
It also helps if the end devices are CCX (Cisco Compatable eXtention) built into their firmware. Most firmware vendors have it built in such as Intel, IBM, etc.. In normal operation a device is sticky, meaning once it has an AP, it tries to stay on it, even if there's a newer AP with better signal nearby. CCX has the smarts to constantly look for a better AP. I'm guessing that the Symbol devices don't have CCX built in.
Actually it really is a mesh. The "controller" is just for management traffic. All Cisco AP's can now use LWAPP, which let's the AP's get all it's management, configurations and control traffic from a centralized device. This means all security configs, polity configs, etc.. are pre-defined on a controller. No pre-configuring each AP before deployment; you just plug in an AP like you would a light bulb and it works.
On the data plane side, they truly are mesh and can talk to any other device in that mesh.
Everything you mentioned sounds fine on paper, but in reality congestion in the cloud rarely happens. If that was the case, Skype, Vonage, Google Talk, and 100's of other VoIP services that travel over the internet wouldn't work, yet they do, right?
There's a bandwidth surplus in the cloud due to overinvestment and most problem happen where you described them, on the last mile/in your equipment. Plus with de-jitter buffers and other mechanisms most VoIP end devices use, losing a few packets once in a while isn't a big deal.
Slashdot Mirror
Ted Turner.
I think you'll start to see this in Large Enterprises over the next few years. With the costs of supporting PC's so high, it would be easier and cheaper to use thin clients/thin wireless tablets that access a grid of virtualized compute resources and a SAN to run their apps.
...back when literacy rates were very low. Suprised someone would try this today.
...If its just an intern program for the younger generation, that great! But from an output perspective, what have we received from it? Is it worth the hype?
topped
I learned it bottom up as - Please Do Not Tell Secret Passwords to Anyone :)
To be really anal:
TCP/UDP Segment > IP Packet > Ethernet Frame/ATM Cell > physical wire
There are plenty of online exchanges for info and ideas, even problems that need solutions that link corporate R&D, Universities, and individual scientists. This is how a lot of companies innovate today - through collaboration, called C&D.
Here's a few:
ninesigma.net
InnoCentive.com
Yet2.com
Modern humans's alone changed a lot over the span of 100,000's of thousands of years and are NOT stable over the span of 2 million years. Same with animals. Using the Galapagos islands as an example, they are estimated to be anywhere from 700,000 to 3.5 million years old. During that time, we can guess that they didn't always have animals on the island, so animals on that islands are considerably less than 2 million years old. During that time, the animals HAVE evlolved and changed from their mainline species. Sequences of animal records on that island show the change in a much shorter time span.
Wrong? You pointed out a 30 year old theory which never caught on with true scientists, yet caught the eyes of creationists because it could show a pattern of species appearing, as if by magic or some intelligent being. Even this article mentions all the controversy surrounding this crackpot theory. Go troll somewhere else.
..that happens over time. We just happen to dig up random fossils and see dramatic changes from the previous, older species. We forget that there were sometimes 10,000's or 100,000's of years in between the two species.
There isn't one "link" between two species. A situation where one day a parent gives birth to a dramatically different, more advanced offspring that is more evolved then the parents doesn't happen. And even if they was a missing link, the chances of that fossil surviving and us finding itwould be near impossible.
I'm not so sure we have too much money in VC's. I agree that we did in the 1990's, but there are far fewer ventures out there today. Most money is socked away in hedge funds these days.
'nuff said
Or they've already maxed out the market share in their current area's of competition. They have 80% market share in Routing and Switching, 60% in wireless, 65% in Linksys, they are now #1 in enterprise voice, #1 in director-class storage, etc... Chambers once said that if they tried to get to 100% market share in Routing and Switching and took all their competitors market cap value, Cisco's market cap price would rise a paltry $25b, nothing investors are going to jump at. Plus it would be a dog fight for little profit or $$. So they need new areas to grow.
Bottom line is that Cisco consolidates networks (consolidated the hodge-podge of networks in the 90's - IPX, DECNet, Banyan Vines, SNA, etc) onto IP and is looking to doing the same today (SAN's, Voice Networks, etc). Analog video is just another network out there that can be put onto IP (imagine all the gear in just one high rise in NYC) and you can see where Cisco's going.
Those all sound like Layer 8 issues (people, support and configs), not HW issues :) Granted HW and software fail on any box from any company, but in my experience most issues are caused by the operators, not the equipment. You can always outsource support of your network if L8 becomes too much of an issue - there are plenty of people out there who know that gear and can do it at a cost effective price; it would be hard to find people to support open source boxes at a cost effective solution...
Can one linux box become a high performance router plus (summary of just new security features released this quarter, never mind all the functionality released the past 3 years)-
Stateful FW Failover
Zone-based Policy Configuration
Cisco Unified Firewall MIB
SSL VPN, including support for Cisco Secure Desktop
(Zone-based Policy Configuration means that administrators will be able to group physical and virtual interfaces into security zones to allow for simplified configuration of firewall rules. Firewall policies can then be applied to a zone rather than an interface. This will also simplify the process of adding or deleting interfaces on a router).
This is just a list of the NEW features released THIS QUARTER:
http://www.cisco.com/univercd/cc/td/doc/product/so ftware/ios124/124newft/124t/124t6/index.htm
* ACL Manageability
* ADSL HWICs
Introduces 2-port ADSL HWIC Hardware.
* ANI Suppression During L2TP Setup
* Certificate - Complete Chain Validation
* Cisco IOS Firewall MIB
* Cisco IOS IPv6 Configuration Library
* Cisco Modem Relay
* Cisco Text Relay for Baudot Text Phones
* Control Plane Logging
* DHCP Option 82 per Interface Support
* DHCP Relay Accounting
* Dynamic Frequency Selection and IEEE 802.11h Transmit Power Control
* Easy VPN Server
* Fax Relay Support for SG3 Fax Machines at G3 Speeds
* FHRP - HSRP Multiple Group Optimization
* Flexible Packet Matching XML Configuration
* In-Service Updates to Gatekeeper Zone Prefix Configuration
* Interface Input Queue Unwedging
* IOS Firewall Stateful Failover
* IP SLAs ICMP Jitter Operation
* IP SLAs--LSP Health Monitor
* IP SLAs RTP-Based VoIP Operation
* Management Plane Protection
* MGCP NAS Package LAPB-TA
* MPLS Embedded Management--LSP Ping for LDP
* MSCHAP Version 2
* NAT ARP Ping
* NAT SCCP Fragmentation Support
* Network Admission Control: Agentless Host Support
* New Voice and Telephony Features in Cisco IOS Releases 12.4T
* OCSP - Server Certification from Alternate Hierarchy
* OER Voice Traffic Optimization
* OSPF Enhanced Traffic Statistics for OSPFv2 and OSPFv3
* OSPF RFC 3623 Graceful Restart Helper Mode
* OSPF: SNMP ifIndex Value for Interface ID in OSPFv2 and OSPFv3 Data Fields
* Packet Mode Services on D Channel
* RIPv2 Monitoring with SNMP Using the RFC 1724 MIB Extensions
* RSVP Agent
* RSVP Application ID Support
* SCCP PLAR with DTMF Ou
The game has long since moved from just forwarding packets to providing intelligence in the network. Now companies want integrated security, voince, application intelligence and application (l5-L7) optimization, QOS, high availability, etc.. none of which you'll find in an open source router. This is why the networking companies stay in business. If companies wanted cheap packet forwarders, they would have bought linksys, 3com, huawiei, hp or any other me-too commodity router. They didn't and Cisco won.
Should ban cordless phones too.
This type of model can be inherently secure as most of the processing, policies and intelligence for these types of applications will be in a central data center. The main difference between this new world and the old is the ability to access the compute resources from anywhere or virtually any device on any network. So if we use a lot of the lessons learned from the "olden days," we can create the best of both worlds.
They have come up with some pretty unsound .com business idea's lately - parallel internet, free wifi for the whole world, etc... I like Google, but it's getting silly.
Steve
Nothing new. From the propaganda side, we've been doing this type of stuff for years - Voice of America, for radio, etc... This is just a new medium. From the disruption of service side, we've also been doing this for years, most recently we debated weather to knock down Al Jazeera.
It also helps if the end devices are CCX (Cisco Compatable eXtention) built into their firmware. Most firmware vendors have it built in such as Intel, IBM, etc.. In normal operation a device is sticky, meaning once it has an AP, it tries to stay on it, even if there's a newer AP with better signal nearby. CCX has the smarts to constantly look for a better AP. I'm guessing that the Symbol devices don't have CCX built in.
Actually it really is a mesh. The "controller" is just for management traffic. All Cisco AP's can now use LWAPP, which let's the AP's get all it's management, configurations and control traffic from a centralized device. This means all security configs, polity configs, etc.. are pre-defined on a controller. No pre-configuring each AP before deployment; you just plug in an AP like you would a light bulb and it works.
On the data plane side, they truly are mesh and can talk to any other device in that mesh.
Everything you mentioned sounds fine on paper, but in reality congestion in the cloud rarely happens. If that was the case, Skype, Vonage, Google Talk, and 100's of other VoIP services that travel over the internet wouldn't work, yet they do, right?
There's a bandwidth surplus in the cloud due to overinvestment and most problem happen where you described them, on the last mile/in your equipment. Plus with de-jitter buffers and other mechanisms most VoIP end devices use, losing a few packets once in a while isn't a big deal.
That theory isn't far-fecthed. We've already done it :)
http://www.gasandoil.com/goc/news/ntr41080.htm