Hey guys.
I really like the RB150 for SOHO and mid range LANs. Does anyone one know for a plan to make a 150 with a higher processor??? (like the 133c to 411 evolution).
Re: Any plan for a Higher Processor in a RB150
My best guess is that the 4xx series will replace 1xx series very soon. Only 1xx thing back that is usefull is 133 and perhaps 150. The 4xx processor has to be the true replacement for these. Then i hope for some N-draft support and a 24 and 48 port version of RB1000Hey guys.
I really like the RB150 for SOHO and mid range LANs. Does anyone one know for a plan to make a 150 with a higher processor??? (like the 133c to 411 evolution).
/Henrik
Re: Any plan for a Higher Processor in a RB150
24 or 48 LANs in a RB1000 would really be awesome...but without higher price 
:D
:D-
-
stephenpatrick
Forum Veteran
- Posts: 702
- Joined:
- Location: UK
- Contact:
Re: Any plan for a Higher Processor in a RB150
... or a nicely modular design, with lots of PCI-express slots, so you can add the ports as you need them 4 or more at a time.
There are plenty of such examples of embedded SBC with PCI expansion slots, but not router-optimised ones with PCI-E which is really needed to avoid PCI-bottleneck at GBE speeds.
Any case, the CPU will be limiting with the ~3Gbps, which for a high-port-count-router may or may not be a problem depending on the application. AFAIK this is the case with most software-driven routers.
High-port-count devices you see like HP Procurve are generally "layer3 switches" which use hardware ASIC to run the protocols at full line rate. Other large routers like Juniper etc have complex core fabric and "network processors" on line cards, which is inherently expensive.
Perhaps a "stacking port" concept could be the answer - put multiple routers in a single enclosure, all wired together with some inter-board high speed bus (PCI-E again?) - and then the software could scale up to allow higher aggregate throughput with the per-board ~3Gbps of processing power per 4 interfaces. Don't know how that would work with the RouterOS architecture or roadmap - but would meet the demands of power users nicely, doesn't add unneccesary cost to a smaller router, and adds some inherent redundancy too in case a single board failed.
Interesting ...
Regards
There are plenty of such examples of embedded SBC with PCI expansion slots, but not router-optimised ones with PCI-E which is really needed to avoid PCI-bottleneck at GBE speeds.
Any case, the CPU will be limiting with the ~3Gbps, which for a high-port-count-router may or may not be a problem depending on the application. AFAIK this is the case with most software-driven routers.
High-port-count devices you see like HP Procurve are generally "layer3 switches" which use hardware ASIC to run the protocols at full line rate. Other large routers like Juniper etc have complex core fabric and "network processors" on line cards, which is inherently expensive.
Perhaps a "stacking port" concept could be the answer - put multiple routers in a single enclosure, all wired together with some inter-board high speed bus (PCI-E again?) - and then the software could scale up to allow higher aggregate throughput with the per-board ~3Gbps of processing power per 4 interfaces. Don't know how that would work with the RouterOS architecture or roadmap - but would meet the demands of power users nicely, doesn't add unneccesary cost to a smaller router, and adds some inherent redundancy too in case a single board failed.
Interesting ...
Regards
Who is online
Users browsing this forum: No registered users and 43 guests