Also, if you get messaging about exporting flows, try changing the port number 
on your zmq endpoint away from 2055 and stick with local host for the IP. 

> On Dec 22, 2015, at 03:06, Yasser Slarmie <[email protected]> wrote:
> 
> Ok, I see lots of output for the netcat dump, meaning that the flow data 
> reaches nprobe. The GUI still doesn't show anything though.
> Nprobe console logging messages for nprobe shows 460 collected packets and 
> 3182 processed flows:
> 
> 
> 22/Dec/2015 11:01:56 [nprobe.c:2476] Flow drop stats:   [0 bytes/0 pkts][0 
> flows]
> 22/Dec/2015 11:01:56 [nprobe.c:2481] Total flow stats:  [0 bytes/0 pkts][0 
> flows/0 pkts sent]
> 22/Dec/2015 11:02:26 [nprobe.c:2589] ---------------------------------
> 22/Dec/2015 11:02:26 [nprobe.c:2590] Average traffic: [0.00 pps][0 b/sec]
> 22/Dec/2015 11:02:26 [nprobe.c:2597] Current traffic: [0.00 pps][0 b/sec]
> 22/Dec/2015 11:02:26 [nprobe.c:2603] Current flow export rate: [0.0 flows/sec]
> 22/Dec/2015 11:02:26 [nprobe.c:2606] Flow drops: [export queue too 
> long=0][too many flows=0]
> 22/Dec/2015 11:02:26 [nprobe.c:2610] Export Queue: 0/512000 [0.0 %]
> 22/Dec/2015 11:02:26 [nprobe.c:2615] Flow Buckets: 
> [active=1][allocated=1][toBeExported=0]
> 22/Dec/2015 11:02:26 [cache.c:1200] Redis Cache [0 total/0.0 get/sec][0 
> total/0.0 set/sec]
> 22/Dec/2015 11:02:26 [nprobe.c:2633] Collector Threads: [460 pkts@0]
> 22/Dec/2015 11:02:26 [nprobe.c:2457] Processed packets: 0 (max bucket search: 
> 0)
> 22/Dec/2015 11:02:26 [nprobe.c:2440] Fragment queue length: 0
> 22/Dec/2015 11:02:26 [nprobe.c:2466] Flow export stats: [0 bytes/0 pkts][0 
> flows/0 pkts sent]
> 22/Dec/2015 11:02:26 [nprobe.c:2473] Flow collection: [collected pkts: 
> 460][processed flows: 3182]
> 22/Dec/2015 11:02:26 [nprobe.c:2476] Flow drop stats:   [0 bytes/0 pkts][0 
> flows]
> 22/Dec/2015 11:02:26 [nprobe.c:2481] Total flow stats:  [0 bytes/0 pkts][0 
> flows/0 pkts sent]
> 
> 
> I also changed the listening IP from localhost to the IP of the server itself 
> as follows:
> 
> ntopng -i tcp://137.158.154.30:2055
> 
> nprobe --zmq tcp://137.158.154.30:2055 --collector-port 9991 -i none -n none 
> -b 1
> 
> 
> What could I check next?
> 
> Regards,
> Yasser
> 
> Date: Tue, 22 Dec 2015 00:56:49 -0600
> From: [email protected]
> To: [email protected]
> Subject: Re: [Ntop-misc] nprobe + ntopng WLAN fields query
> 
> Awesome. You are not running into the same issue that I did which is good. I 
> am also a bit fuzzy on the PF_Ring interaction as well, but I have it enabled 
> also and still got bit by rpf.....
> 
> I see you ran netcat on port 2055, I think we got our wires crossed on what 
> port you were using. I see from the tcpdump that you are using 9991 for flow 
> data.
> 
> What do you see if you stop nprobe/ntopng, run 'netcat -lu 9991 | hexdump -C' 
> ? This should confirm that the flow-data is actually making it through any 
> filtering such as iptables/ufw.
> 
> Erik
> 
> On Tue, Dec 22, 2015 at 12:33 AM, Yasser Slarmie <[email protected]> 
> wrote:
> Hi Erik,
> 
> The server has only 1 nic and a single default-gateway and there is IP 
> reachability between the netflow exporter and collector. I turned rpf off, 
> just for the sake of argument, but I still don't see the packets passing the 
> kernel. 
> 
> PF_Ring is enabled, but I don't know enough about it to say whether the 
> packets will bypass the kernel or not..
> 
> ~$ netstat -rn
> Kernel IP routing table
> Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
> 0.0.0.0         137.158.154.1   0.0.0.0         UG        0 0          0 eth0
> 137.158.154.0   0.0.0.0         255.255.255.0   U         0 0          0 eth0
> 
> 
> 
> root@devubunfl001:~# cat /proc/sys/net/ipv4/conf/eth0/rp_filter
> 1
> root@devubunfl001:~# echo 0 > /proc/sys/net/ipv4/conf/eth0/rp_filter
> root@devubunfl001:~# cat /proc/sys/net/ipv4/conf/eth0/rp_filter
> 0
> 
> 
> root@devubunfl001:~# netcat -lU 2055 | hexdump -C
> ^C
> 
> 
> 
> root@devubunfl001:~# ufw status
> Status: active
> 
> To                         Action      From
> --                         ------      ----
> 9991/udp                   ALLOW       Anywhere
> 22                         ALLOW       Anywhere
> 80/tcp                     ALLOW       Anywhere
> 123                        ALLOW       Anywhere
> 3000/tcp                   ALLOW       Anywhere
> 2055/tcp                   ALLOW       Anywhere
> 9991/udp (v6)              ALLOW       Anywhere (v6)
> 22 (v6)                    ALLOW       Anywhere (v6)
> 80/tcp (v6)                ALLOW       Anywhere (v6)
> 123 (v6)                   ALLOW       Anywhere (v6)
> 3000/tcp (v6)              ALLOW       Anywhere (v6)
> 2055/tcp (v6)              ALLOW       Anywhere (v6)
> 
> 
> 
> root@devubunfl001:~# tcpdump -n host 137.158.248.10
> tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
> listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes
> 08:15:53.230349 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 
> 1320
> 08:15:55.232036 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 816
> 08:16:08.229221 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 456
> 08:16:11.229323 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 744
> 08:16:12.228999 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 168
> 08:16:13.229034 IP 137.158.248.10.32773 > 137.158.154.30.9991: UDP, length 456
> ^C
> 6 packets captured
> 7 packets received by filter
> 0 packets dropped by kernel
> 
> 
> 
> Date: Mon, 21 Dec 2015 16:09:52 -0600
> From: [email protected]
> 
> To: [email protected]
> Subject: Re: [Ntop-misc] nprobe + ntopng WLAN fields query
> 
> Exactly what I am referring to. Your plight sounds exactly like mine as I 
> just got bit by this and wasted Luca and crew's time sorting it out. Ubuntu 
> runs RPF check in the kernel and if the Netflow traffic arrives on an 
> interface that doesn't have a route back to the source via the interface it 
> was received on, the kernel rejects it. 
> 
> You can verify by running netcat -lu 2055 | hexdump -C after starting the 
> netflow feed from the router.
> 
> Tcpdump runs at the interface pre RPF check, Netcat will bind to the port and 
> tell you if the kernel passed the traffic. If you do not see traffic data, 
> it's being dropped before nProbe can capture it.
> 
>   
> 
> On Mon, Dec 21, 2015 at 3:54 PM, Yasser Slarmie <[email protected]> 
> wrote:
> Ntop and nprobe are on the same server. Are you perhaps alluding to urpf that 
> may be breaking things? The netflow packets arrive on the server, so I 
> believe nprobe should be able to interpret it without any routing back to the 
> source or requiring to do so.
> 
> Sent from my Windows Phone
> From: Erik Schmersal
> Sent: ‎2015-‎12-‎21 07:37 PM
> 
> To: [email protected]
> Subject: Re: [Ntop-misc] nprobe + ntopng WLAN fields query
> 
> Is netflow being received on the same interface where Ubuntu's default route 
> pointing out of? Or is the route back to the flow source pointing out the 
> same interface that the flows are coming in on?
> 
> On Mon, Dec 21, 2015 at 9:24 AM, Yasser Slarmie <[email protected]> 
> wrote:
> Hi Eric,
> 
> My commands are:
> 
> ntopng -i tcp://127.0.0.1:2055 &
> 
> nprobe --zmq "tcp://127.0.0.1:2055" --collector-port 9991 -i none -n none -b 
> 1 &
> 
> Regards,
> Yasser
> 
> From: [email protected]
> Date: Mon, 21 Dec 2015 07:59:00 -0600
> To: [email protected]
> Subject: Re: [Ntop-misc] nprobe + ntopng WLAN fields query
> 
> 
> Can you post your nProbe and nTop commands?
> 
> On Dec 21, 2015, at 03:29, Yasser Slarmie <[email protected]> wrote:
> 
> Hello guys,
> 
> I don't know how to bottom-post on a thread from April 2015, but the above 
> subject line is still the same.
> 
> I'm implementing ntopng and nprobe for a University (so professional license 
> is installed and working) and they want to test it specifically for netflow 
> exports coming from their 6 Cisco Wireless LAN Controllers.
> 
> I have ntopng and nprobe setup but the GUI doesn't interpret any of the 
> received netflow data. The packets do arrive on the Ubuntu server though. As 
> a test, I exported traffic from their Cisco 6500 switch, and the  GUI 
> displays the data correctly.
> 
> I took a pcap dump of the received WLC traffic. It's available here: 
> http://1drv.ms/1mf2iuC
> 
> Could someone please help with what I should do to get the data to display?
> 
> Kind regards,
> Yasser
> 
> 
> _______________________________________________
> 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

Reply via email to