Yep, pretty clear now ... Many thanks Alfredo.
On Tue, Jul 7, 2015 at 1:36 PM, Alfredo Cardigliano <[email protected]> wrote: > As reported in examples_zc/README.kvm, you do not need to use virtio nics, > you need to run zbalance_ipc on the host to balance traffic to SPSC queues, > and attach your IDSs to those queues (ZC will do the magic). > > Alfredo > >> On 07 Jul 2015, at 15:32, C. L. Martinez <[email protected]> wrote: >> >> Ok, perfect .. I will wait to this new version. >> >> But, I don't see how can I balance traffic using ZC using libvirt/qemu: >> >> a) Configure two virtio nics in every guest: one nic for management >> and another to sniff traffic. >> b) Start every guest with a monitor socket (two sockets in total from >> kvm host side, correct?). >> >> >> Is this correct?? If it is, how traffic is balanced across these vms ?? >> >> >> >> On Tue, Jul 7, 2015 at 1:24 PM, Alfredo Cardigliano >> <[email protected]> wrote: >>> >>>> On 07 Jul 2015, at 15:17, C. L. Martinez <[email protected]> wrote: >>>> >>>> Many thanks Alfredo. >>>> >>>> Some questions: >>>> >>>> a) Do I need to patch qemu released by vendor?? I would like to avoid >>>> this if it is possible… >>> >>> Last version we tested required that due to a bug, we will provide a new >>> tested version (which likely fixed that) ASAP. >>> >>>> b) If i am not wrong, I need to cconfigure pf_ring module with the >>>> following option: transparent_mode=0. ZC will be available using this >>>> mode?? >>> >>> ZC just ignores kernel flags. BTW transparent_mode is no more needed in any >>> case. >>> >>> Alfredo >>> >>>> >>>> >>>> >>>> >>>> On Tue, Jul 7, 2015 at 1:09 PM, Alfredo Cardigliano >>>> <[email protected]> wrote: >>>>> Of so you need to replicate the same traffic form the same interface to 3 >>>>> VM, got it. >>>>> Please note you cannot use ZC drivers because your card is not supported >>>>> (Intel only at the moment), >>>>> however you can still use ZC to distribute traffic across VMs. In this >>>>> case you need pf_ring on both >>>>> host and guest systems, please have a look at >>>>> https://github.com/ntop/PF_RING/blob/dev/userland/examples_zc/README.kvm >>>>> >>>>> Alfredo >>>>> >>>>>> On 07 Jul 2015, at 14:59, C. L. Martinez <[email protected]> wrote: >>>>>> >>>>>> Yes, but not via pci-passthorugh ... I need to pass the same phys nic >>>>>> to three vms ... >>>>>> >>>>>> On Tue, Jul 7, 2015 at 12:56 PM, Alfredo Cardigliano >>>>>> <[email protected]> wrote: >>>>>>> Hi >>>>>>> it depends on your architecture, do you have to bind an interface to a >>>>>>> single VM? >>>>>>> >>>>>>> Alfredo >>>>>>> >>>>>>>> On 07 Jul 2015, at 14:50, C. L. Martinez <[email protected]> wrote: >>>>>>>> >>>>>>>> Hi all, >>>>>>>> >>>>>>>> I need to use pf_ring to setup some virtual IDS host sensors in >>>>>>>> CentOS 6.x/7.x KVM hosts. >>>>>>>> >>>>>>>> With PF_RING recent releases, do I need to install it in both (kvm >>>>>>>> host and kvm guests) or only in kvm guest?? >>>>>>>> >>>>>>>> These kvm hosts have broadcom interfaces: >>>>>>>> >>>>>>>> tg3.c:v3.137 (May 11, 2014) >>>>>>>> tg3 0000:03:04.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 >>>>>>>> tg3 0000:03:04.0: eth1: Tigon3 [partno(011276-001) rev 9003] >>>>>>>> (PCIX:133MHz:64-bit) MAC address 80:c1:6e:62:f4:44 >>>>>>>> tg3 0000:03:04.0: eth1: attached PHY is 5714 (10/100/1000Base-T >>>>>>>> Ethernet) (WireSpeed[1], EEE[0]) >>>>>>>> tg3 0000:03:04.0: eth1: RXcsums[1] LinkChgREG[0] MIirq[0] ASF[1] >>>>>>>> TSOcap[1] >>>>>>>> tg3 0000:03:04.0: eth1: dma_rwctrl[76148000] dma_mask[40-bit] >>>>>>>> tg3 0000:03:04.1: PCI INT B -> GSI 17 (level, low) -> IRQ 17 >>>>>>>> tg3 0000:03:04.1: eth2: Tigon3 [partno(011276-001) rev 9003] >>>>>>>> (PCIX:133MHz:64-bit) MAC address 80:c1:6e:62:f4:45 >>>>>>>> tg3 0000:03:04.1: eth2: attached PHY is 5714 (10/100/1000Base-T >>>>>>>> Ethernet) (WireSpeed[1], EEE[0]) >>>>>>>> tg3 0000:03:04.1: eth2: RXcsums[1] LinkChgREG[0] MIirq[0] ASF[0] >>>>>>>> TSOcap[1] >>>>>>>> tg3 0000:03:04.1: eth2: dma_rwctrl[76148000] dma_mask[40-bit] >>>>>>>> tg3 0000:03:04.1: irq 61 for MSI/MSI-X >>>>>>>> tg3 0000:03:04.1: eth1: Link is up at 1000 Mbps, full duplex >>>>>>>> tg3 0000:03:04.1: eth1: Flow control is on for TX and on for RX >>>>>>>> >>>>>>>> Thanks. >>>>>>>> _______________________________________________ >>>>>>>> Ntop-misc mailing list >>>>>>>> [email protected] >>>>>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> Ntop-misc mailing list >>>>>>> [email protected] >>>>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >>>>>> _______________________________________________ >>>>>> Ntop-misc mailing list >>>>>> [email protected] >>>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >>>>> >>>>> >>>>> _______________________________________________ >>>>> Ntop-misc mailing list >>>>> [email protected] >>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >>>> _______________________________________________ >>>> Ntop-misc mailing list >>>> [email protected] >>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >>> >>> >>> _______________________________________________ >>> Ntop-misc mailing list >>> [email protected] >>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc >> _______________________________________________ >> Ntop-misc mailing list >> [email protected] >> http://listgateway.unipi.it/mailman/listinfo/ntop-misc > > > _______________________________________________ > Ntop-misc mailing list > [email protected] > http://listgateway.unipi.it/mailman/listinfo/ntop-misc _______________________________________________ Ntop-misc mailing list [email protected] http://listgateway.unipi.it/mailman/listinfo/ntop-misc
