We recently upgraded a number of RB1100 from 5.21 to 6.7 and are now seeing spikes (!?) in both the real time byte/packet graphs and in the Tx/RX columns in Interfaces. They appear to be real because they seen by other devices upstream/downstream. The heavier the traffic, the bigger the spike. They are about double the nominal traffic level, a 40M traffic level will show spikes to 80M, etc. The packets in the packet graph double as well. Interval is around 45 seconds more or less. This can’t be good…
I can confirm.
We have the same situation. Spikes are at every 35 secs. 6.8rc1 does’t repair a problem. We have this issue on RB1100Hx2, RB 1100 and RB2011.
I can read before on any Polish forum that users have similar problems.
It surprise me that so few people can see it or why don’t they write.
Lastly I checked on RB750GL and spikes are same. Before I wrote here I issued a ticket in my upstream provider because I thought the spikes are from WAN. They don’t see any spikes at his interface. Now I am sure there is a problem with ROS.
One thing I see the spikes are only on GBit ports.
