We have a client that have several ME-3400E switches, their problem was when they are using their Ethernet tester (http://www.jdsu.com/en-us/Test-and-Measurement/Products/a-z-product-list/Pages/smartclass-ethernet.aspx) in compliance to RFC2544, packets are dropped when their tester is set to 802.3, whereas if they set it to DIX (Ethernet or Ethernet II) packets flows smoothly.
My question is why is packets being dropped 802.3 is set? Do we have to adjust some configuration on the ME-3400E switches for 802.3 to work?
we actually all use Ethernet II encapsulation, and this has been become part of IEE 802.3 in last revision of standard
so the question may be: is the instrument referrring to original older IEEE 802.3 encapsulation that is different?
if so the bad results can be explained.
>> Do we have to adjust some configuration on the ME-3400E switches for 802.3 to work?
you don't need to do anything if your customer is going to send/receive IPv4 and IPv6 traffic they will use Ethernet II encapsulation.
Hope to help
Thanks for your reply, I read the link your provided and also read some other documents as well. Base on my understanding there are 4 types of ethernet encapsulation and Ethernet II being the oldest?
802.3 LLC + SNAP'
- Can you further exlplain to me when you said the "We all use Ethernet II encapsulation".
- Also where in the 802.3 did Ethernet II become part of it? Is it in 802.3 LLC + SNAP? If it is then we use the 802.3 LLC + SNAP encapsulation then this is the same as using Ethernet II encapsulation?
- Lastly of all the four encapsulation types, which is the best, widely used today specially by the ISP providers?
Ethernet II should have been incorporated in IEEE 802.3 1998 revision because it was already the most used as a variant, not as part of the other standardized encapsulations.
you can easily demonstrate this to yourself if you run wireshark or ethereal on your PC
or you can use wikipedia
or you can go to wireshark sample captures
after having installed wireshark you can open the capture file and you can see
Ethernet II as the encapsulation
I have taken the first packet of the above example capture and after having exported it in a txt file here it is the contents:
No. Time Source Destination Protocol Info
1 0.000000 192.168.0.33 192.168.0.15 TCP bgp > gtp-control [FIN, PSH, ACK] Seq=1 Ack=1 Win=16101 Len=0
Frame 1 (60 bytes on wire, 60 bytes captured)
>>>Ethernet II, Src: Cisco_35:0e:1c (00:00:0c:35:0e:1c), Dst: DellComp_23:c5:95 (00:c0:4f:23:c5:95)
Destination: DellComp_23:c5:95 (00:c0:4f:23:c5:95)
Source: Cisco_35:0e:1c (00:00:0c:35:0e:1c)
Type: IP (0x0800)
Internet Protocol, Src: 192.168.0.33 (192.168.0.33), Dst: 192.168.0.15 (192.168.0.15)
Transmission Control Protocol, Src Port: bgp (179), Dst Port: gtp-control (2123), Seq: 1, Ack: 1, Len: 0
Thanks for sharing your thoughts to me, also for the wireshark capture, I also used wireshark and recall that there is Ethernet II encapsulation in the packets, somehow manage to understand Ethernet II frame.
Two last question:
1. Is that is it safe to say that the reason why packets are dropped when they flood the ME-3400E with 802.3 frame is because Ethernet II is not yet incorporated to it? And that is normal and they we should use Ethernet II as the source frame when using stress test on ME-3400E switches?
2. Is it normal for the ME-3400E to drop packets when they are encapsulated with 802.3 frame or do they need 802.3 frame to pass thru? If so, how can I let the ME-3400E to accept 802.3 frame?
Thanks you very much
Answers to your last questions:
1. No, i think you've misunderstood the issue here. There is no problem with any incorporation of anything anywhere. It's just history.
2. Good reasoning is to stress test technologies, that are going to be used. You need to test Ethernet|| because it's pretty much all the encapsulation you'll see. Why would you spend time stress testing hypothetical scenarios? You just won't see nearly any LLC, apart from the specifics, like CDP.
P.s. me3400 may process switch some of the specifics.
I saw in wireshark the frames with 802.3 encapsulation for PVST, as you said also in CDP,.. When using the tester from my first post above, frames with 802.3 encapsulations are being dropped on ME-3400E switch, while Ethernet II frames passess thru... Is it an Ok scenario? Since our client is a small branch of service provider, do (service provider) need 802.3 frames to pass thru them or only Ethernet II frames?
If they SP also need 802.3 frames to pass thru them aside from Ethernet II frames, do you know the reason why they are being dropped on ME-3400E switches?
everything is OK in this network as IPv4 packets are carried within Ethernet II frames
>> If they SP also need 802.3 frames
if the ME 3400 performs routing of IPv4 packets and/or IPv6 packets to/from central site, you need only Ethernet II
if you already use wireshark you should have realized that what is not IP is not important anymore.
STP BPDUs and CDP frames are locally processed and not forwarded to any other device.
Hope to help
Thank you very much Giuseppe, your a big help on my querry about this. I think I almost understand your explanation and you are correct that IPv4/v6 packets will be routed to/from the CO site.
I'm just a little confuse as to why they said that 802.3 frames are being dropped when they use their tester device, as they said that they need this because they are in compliance of rfc 2544. Do you happened to know why ME3400E drops 802.3 frames? Any solution/resolution for ME3400E not to droped 802.3 frames or it is the behavior of the switch itself?