There has been a steady stream of complaints about flaky operation of some draft 11n routers, most notably the WRT610N. So I have begun to do four and eight hour stress tests of some products.
The test setup for the WRT400N was similar to the one I used for the short-term stress testing of the WRT610N. I basically set up two draft 11n clients, one associated to each radio. I then ran an IxChariot throughput test blasting full-speed bi-directional traffic between each wireless client and a single wired one. This time, however, I did not also simultaneously run traffic through the router.
Figure 11 shows that the test ran successfully for the full eight hours with no disconnects. There obviously was a wide throughput spread. But expanding out the time axis shows this was from periodic throughput "dropouts", which I didn't see during the shorter one minute tests.
Figure 11: Eight hour wireless stress test
I'll note that the 400N got a bit warm during the test. But apparently not hot enough to cause any problems. I suspect the absence of a gigabit switch helped to keep things cool.
Cisco seems to believe in simultaneous dual-band routers, since the WRT400N is their third run at it. And it appears that they think you shouldn't have to pay a lot for the benefit, since the 400N is around $50 cheaper than its flagship WRT610N.
They say that it's better to cannibalize your own products than have someone else do it for you. From what I can see, the WRT400N is going to do just that to the 610N. But not without causing some soul-searching for those who believe that it's crazy to buy a draft 11n router without a gigabit switch.
Draft 802.11n routers won't replace a 100 Mbps Ethernet connection for either speed or reliability and the WRT400N is no exception. But it seems to be a well-priced, well-designed addition to a product category filled with promise-long and deliver-short offerings. It's worthy of your consideration, if not your hard-earned cash.