no Force10?By Anonymous on January 19, 2010, 11:47 pmI thought Force10 was the dominant player in this market. Or at least well known for its 10G capabilities within the data center. Any opinions as to why it was not included in this test?
ThroughputBy Anonymous on March 16, 2010, 5:28 pmWhy do we not have the throughput results mentioned here? Are we to assume all vendors were able to perform 100% zero packet loss performance?
Throughput??By Anonymous on March 16, 2010, 5:26 pmWhy do we not see the throughput results? Are we to assume alll vendors were able to achieve 100% zero packet loss?
Cisco shifting focusBy Anonymous on March 2, 2010, 10:08 amCISCO is now busy in acquiring companies in order to expand their product portfolio. As a result they are losing focus on networking - switching and routing in particular!!
latencyBy Anonymous on January 22, 2010, 3:28 pmI work at a certain type of firm in NYC where latency is our #1 priority. Having done similar benchmark testing between these products, I think something that is missing from this test is the switch performance under extremely bursty conditions (microbursts). Chances are that anyone with a business reason to care about latency at this level will also care about microbursts. During our testing, we saw that the Nexus was able to handle these conditions more effectively than the competitors. Plus, we just felt safer going with the market leader in such a critical part of our network.
Safety with MediocrityBy Anonymous on January 22, 2010, 10:22 pmNobody gets fired for buying Cisco, but increasingly budget and latency is scrutinized
So go ahead and pay more money and get less or look at alternatives
Handcrafting microburset data does not change the simple fact that Arista latency is always lower than Nexus 5k. Looks like Cisco has a multicast issues too.Maybe a bug but even if its fixed the real world scalability is a challenge
Latency and Microburst?By Anonymous on January 22, 2010, 8:12 pmThe test data that Cisco shares about microburst is hand crafted to fit in just in the Nexus buffers and overrun everyone elses. Add one single bit more of data and they will do just as bad as everyone else (fan-in is fan-in!). The penalty for the increased buffer size on the Nexus 5000 is higher latency. The Arista and BNT clock in at less then 1us and the Nexus is 3.2us. If your number one issues is latency this is a no brainer. You can design your network and data flows so that the buffering is never an issue but you can never make the Nexus faster.
Impressive Arista!By Anonymous on January 21, 2010, 11:40 pmWhy dont you re-run the test with Arista supports FCoE natively (I think you missed the whole point of the Nexus)
BrocadeBy Anonymous on January 21, 2010, 2:30 pmThe article says the Cisco Nexus is the only switch manufacturer that delivers converged FCoE. This is totally false. Brocade also delivers this capability and does it better than Cisco.
Impressive Arista!By Anonymous on January 20, 2010, 11:18 amI thought HP and Cisco were switching leaders? They come last?
Looks like the 10GE market has new innovative players with best of breed price-performance-capabilities
Congratulations Arista Networks!
MIA Foundry,Force 10, Juniper
I thought HP and Cisco were switching leadersBy Anonymous on January 23, 2010, 11:24 amNo obviously they are not cutting edge technology. That's the reason why HP acquired 3Com - H3C. That's cutting edge technology....
Test AreasBy Anonymous on January 20, 2010, 1:25 amIt would have been really a data center testing if more data center related areas would have been tested, for eg support to virtualization, interoperability with Servers, Latency, througput, redundancy etc.
Performance vs FeaturesBy Anonymous on January 23, 2010, 4:34 pmThe test really emphasized key performance metrics and also Layer 2 features and pretty exhaustive suite of tests!
Enhanced capabilities like virtualization, issu, layer 3,virtualization and storage were discussed but not the top criteria. Had they been so Cisco Nexus 5000 would have underperformed again as they have no Layer 3 and little issu
Test AreasBy david_newman on January 21, 2010, 2:20 amLatency, jitter and throughput were key metrics in this test. I agree that virtualization support would be an excellent addition (we covered it this time in features, but am planning to include performance measurements in future tests).
How about is Enterasys?By Anonymous on January 20, 2010, 1:59 amI think Enterasys is also putting 10G switch for data center on show. Why it was not included in this?
Force10By Anonymous on January 20, 2010, 1:43 amIn the Article, Force10 was invited but declined to participate.. They have a 24 port 10 Gig switch but it does not run the Force10 Operating System...
The Connected Enterprise
no Force10?By Anonymous on January 19, 2010, 11:47 pmI thought Force10 was the dominant player in this market. Or at least well known for its 10G capabilities within the data center. Any opinions as to why it was not included in this test?
shijinBy Anonymous on April 12, 2010, 2:13 am fake handbags | replica handbag | designer handbags | fake designer handbags | replica handbag | replica bag | replica handbags | Givenchy Replica handbag | Versace bags | Burberry bags | quality replica Jimmy Choo bags | | Fake MiuMiu bag | Balenciaga bag
ThroughputBy Anonymous on March 16, 2010, 5:28 pmWhy do we not have the throughput results mentioned here? Are we to assume all vendors were able to perform 100% zero packet loss performance?
Throughput??By Anonymous on March 16, 2010, 5:26 pmWhy do we not see the throughput results? Are we to assume alll vendors were able to achieve 100% zero packet loss?
Cisco shifting focusBy Anonymous on March 2, 2010, 10:08 amCISCO is now busy in acquiring companies in order to expand their product portfolio. As a result they are losing focus on networking - switching and routing in particular!!
What about Extreme Networks?By Anonymous on February 20, 2010, 10:39 am How does it compare with Extreme's switches? http://www.extremenetworks.com/
cheap god shoesBy Anonymous on February 1, 2010, 10:57 pmcheap fake watches
oris watches
Burberry replica
Rolex Masterpiece watches
Piaget replica
nike trainers
latencyBy Anonymous on January 22, 2010, 3:28 pmI work at a certain type of firm in NYC where latency is our #1 priority. Having done similar benchmark testing between these products, I think something that is missing from this test is the switch performance under extremely bursty conditions (microbursts). Chances are that anyone with a business reason to care about latency at this level will also care about microbursts. During our testing, we saw that the Nexus was able to handle these conditions more effectively than the competitors. Plus, we just felt safer going with the market leader in such a critical part of our network.
Safety with MediocrityBy Anonymous on January 22, 2010, 10:22 pmNobody gets fired for buying Cisco, but increasingly budget and latency is scrutinized So go ahead and pay more money and get less or look at alternatives Handcrafting microburset data does not change the simple fact that Arista latency is always lower than Nexus 5k. Looks like Cisco has a multicast issues too.Maybe a bug but even if its fixed the real world scalability is a challenge
Latency and Microburst?By Anonymous on January 22, 2010, 8:12 pmThe test data that Cisco shares about microburst is hand crafted to fit in just in the Nexus buffers and overrun everyone elses. Add one single bit more of data and they will do just as bad as everyone else (fan-in is fan-in!). The penalty for the increased buffer size on the Nexus 5000 is higher latency. The Arista and BNT clock in at less then 1us and the Nexus is 3.2us. If your number one issues is latency this is a no brainer. You can design your network and data flows so that the buffering is never an issue but you can never make the Nexus faster.