vpoluektov
New Around Here
If any of you have been playing with third-party firmware for RT-N16, please share your experiences.
My priorities for the home network are:
Essential (if anything from this list is missing, I will not be able to use the firmware):
Complete reliability of wired segment and internet connection (dynamic IP, can be set to static as it's static anyway - only DNS servers change occasionally). I understand 0% packet loss due to the faults in the router is unrealistic, but the target should be at least 0.1%.
Complete reliability of wireless segment in mixed g/n mode. I may use either the stock antennas or WL-ANT150G from Asus (+5 dB) with the router; the most important client is using PCE-N13 from Asus with the same WL-ANT150G, with no option to use stock antennas, as the machine is built in Silverstone RV02 case, and the metal grill on top kills the signal. The rest of the clients are notebooks with various Intel chip with no 802.11n. The 2.4 GHz environment is pretty noisy, about 15 APs visible, 4 of them with signal strength over -70 dB.
Being able to forward specific ports to specific IP addresses without using DMZ. RT-N16 had a problem with that with 3.0.0.3.78 stock firmware; ports like TCP/22 and TCP/3389, which are tested for vulnerabilities every minute or so, would stay forwarded for an hour at most, unless a DMZ host is used. uTorrent port would stay forwarded forever, and it gets a lot more connection, so this must be some sort of IDS in action; unfortunately, I need to forward these ports to two different addresses, so DMZ is not an option.
Useful, but not essential:
Decent 802.11n performance. Again, another RN-N16 with WL-ANT150 gives well over 100 Mbit 802.11n bandwidth on 3.0.0.3.78 - but with wireless connection drop for about 2 seconds every hour, which is not acceptable to me. I don't need 100 Mbit, but would like at least 60. If "decent performance" and "stable connection" cannot work together, I wouldn't mind switching router settings between these two modes when I need performance, which I don't often do - but only as long as this doesn't require the reboot of the router. Restart of the wireless interface would clearly be necessary, and is not a problem.
Working QoS; upstream only, of course. There will be a single host generating continuous outgoing traffic, and it would be nice if it didn't interfere with anything.
Being able to cope with 1 Mbit upstream traffic 24x7 (2 Mbit if QoS is working).
Completely unnecessary:
USB functionality, VPN connectivity from the outside, download client on the router, firewalling traffic from the inside, uPnP, media servers... basically anything that's not mentioned above.
Having done some research, I have seen that it is often said that dd-wrt is not stable in mixed g/n mode; original TomatoUSB is stable, but is pretty long in the tooth now, so the stock firmware may have pulled ahead since the end of 2010; the current build of Toastman's Tomato is said to be unstable wirelessly, and searching through the previous versions with 30-30-30 NVRAM erase before and after each flash is going to be painful; OpenWRT does not even officially support RT-N16 yet, and while there are beta builds floating around that may be beta than nothing, I doubt they are better than stock one; stock firmware would look like a perfect solution, if not for those 2 second connection drops (on all wireless clients at the same time, so this is the router/firmware problem, not client problem) on 3.0.0.3.78. So any first-hand experience with this router in the similar conditions would be very welcome.
My priorities for the home network are:
Essential (if anything from this list is missing, I will not be able to use the firmware):
Complete reliability of wired segment and internet connection (dynamic IP, can be set to static as it's static anyway - only DNS servers change occasionally). I understand 0% packet loss due to the faults in the router is unrealistic, but the target should be at least 0.1%.
Complete reliability of wireless segment in mixed g/n mode. I may use either the stock antennas or WL-ANT150G from Asus (+5 dB) with the router; the most important client is using PCE-N13 from Asus with the same WL-ANT150G, with no option to use stock antennas, as the machine is built in Silverstone RV02 case, and the metal grill on top kills the signal. The rest of the clients are notebooks with various Intel chip with no 802.11n. The 2.4 GHz environment is pretty noisy, about 15 APs visible, 4 of them with signal strength over -70 dB.
Being able to forward specific ports to specific IP addresses without using DMZ. RT-N16 had a problem with that with 3.0.0.3.78 stock firmware; ports like TCP/22 and TCP/3389, which are tested for vulnerabilities every minute or so, would stay forwarded for an hour at most, unless a DMZ host is used. uTorrent port would stay forwarded forever, and it gets a lot more connection, so this must be some sort of IDS in action; unfortunately, I need to forward these ports to two different addresses, so DMZ is not an option.
Useful, but not essential:
Decent 802.11n performance. Again, another RN-N16 with WL-ANT150 gives well over 100 Mbit 802.11n bandwidth on 3.0.0.3.78 - but with wireless connection drop for about 2 seconds every hour, which is not acceptable to me. I don't need 100 Mbit, but would like at least 60. If "decent performance" and "stable connection" cannot work together, I wouldn't mind switching router settings between these two modes when I need performance, which I don't often do - but only as long as this doesn't require the reboot of the router. Restart of the wireless interface would clearly be necessary, and is not a problem.
Working QoS; upstream only, of course. There will be a single host generating continuous outgoing traffic, and it would be nice if it didn't interfere with anything.
Being able to cope with 1 Mbit upstream traffic 24x7 (2 Mbit if QoS is working).
Completely unnecessary:
USB functionality, VPN connectivity from the outside, download client on the router, firewalling traffic from the inside, uPnP, media servers... basically anything that's not mentioned above.
Having done some research, I have seen that it is often said that dd-wrt is not stable in mixed g/n mode; original TomatoUSB is stable, but is pretty long in the tooth now, so the stock firmware may have pulled ahead since the end of 2010; the current build of Toastman's Tomato is said to be unstable wirelessly, and searching through the previous versions with 30-30-30 NVRAM erase before and after each flash is going to be painful; OpenWRT does not even officially support RT-N16 yet, and while there are beta builds floating around that may be beta than nothing, I doubt they are better than stock one; stock firmware would look like a perfect solution, if not for those 2 second connection drops (on all wireless clients at the same time, so this is the router/firmware problem, not client problem) on 3.0.0.3.78. So any first-hand experience with this router in the similar conditions would be very welcome.
