Maurizio it's hard to give you a definitive answer. One of the reference tests I do is the following - Core2Duo 1.86 GHz PC, 2 GB RAM - PF_RING 4.6.x - Intel 1G adapter with PF_RING-aware driver (no TNAPI) - Ixia injecting 200k flows simultaneously (64 bytes packets) - nprobe 6.3.x (nprobe -O 1 -i ethX, no other flags, netFlow v5)
nprobe can handle with no loss, converting packets to NetFlow, ~ 1.12 Mpps (incoming traffic rate with no packet drops) Luca On Mar 28, 2011, at 2:42 PM, Maurizio Molina wrote: > Hi, > has anybody on this list performed stress tests with nprobe? (for the time > being, I'm interested in testing the netflow exporting capability only). > If yes, I'd appreciate information about: > - the scenario you used for generating the load (HW, SW) > - what (if anything) "put in trouble" nprobe (packet load? flow load? weird > TCP traffic like half-open connections, protocol used to export NetFlow > packets, etc.) > - performance figures (if available) > > Thanks, > Maurizio Molina > _______________________________________________ > Ntop-misc mailing list > [email protected] > http://listgateway.unipi.it/mailman/listinfo/ntop-misc --- If you can not measure it, you can not improve it - Lord Kelvin _______________________________________________ Ntop-misc mailing list [email protected] http://listgateway.unipi.it/mailman/listinfo/ntop-misc
