hi @ all,

i am for the firt time using MT devices in a gigabit network or at least try to do so. with plain ip routing without any fancy stuff from ether1 to ether2 on the RB600, connection tracking off and CPU clock at max i get no more then 276 mbit/s TCP throughput. i test with as single wget session to /dev/null to make sure file I/O does not screw up my results. on the uplink to the core router i get over 950 mbit/s. if i connect ether1 instead of my test client and connect my test client to ether2, which is another routed network, i only get 276 mbit/s max.

have others made similar experiences ? just like to see what is considered 'normal'.

i am using ROS 5.2

thanks
Jan

[edit]
support file is send to MT support btw.
[/edit]

You should try downgrading to ROS 4.16 or 4.17. I can easily reach 500 Mbit/s using VPLS transport (which is arguably more stressful) between two RB 600s...

All I can tell you, You are not alone. There is definitively wrong something in RouterOS 5.2. I would be very happy if I could reach that throughput. Could you test in on WAN? Incoming traffic is acceptable for me, but outgoing is terrible. Downgrading to 4.17 helps much, but not as much I would like.

hi @ all,

i am for the firt time using MT devices in a gigabit network or at least try to do so. with plain ip routing without any fancy stuff from ether1 to ether2 on the RB600, connection tracking off and CPU clock at max i get no more then 276 mbit/s TCP throughput. i test with as single wget session to /dev/null to make sure file I/O does not screw up my results. on the uplink to the core router i get over 950 mbit/s. if i connect ether1 instead of my test client and connect my test client to ether2, which is another routed network, i only get 276 mbit/s max.

have others made similar experiences ? just like to see what is considered 'normal'.

i am using ROS 5.2

thanks
Jan

[edit]
support file is send to MT support btw.
[/edit]

hi,

downgrading to 4.17 did not help. the RB also has a hotspot on an R52n which is not used (no clients connected) during the test. just to clarify, i am testing routed between ether1 and ether2. if i disable the unused radio i get roughly double the speed !!! why the heck is that ? this interface is NOT used at all ?

regards
Jan

Was this ever resolved?