View entire discussion ( 16 comments) More posts from the Cisco community Continue browsing in rCisco rCisco Welcome to Cisco 46.1k Members 131 Online Created Jun 18, 2008 Join help Reddit App Reddit coins Reddit premium Reddit gifts about careers press advertise blog Terms Content policy Privacy policy Mod policy Reddit Inc 2020.Press question mark to learn the rest of the keyboard shortcuts Log in sign up User account menu 2 No packet loss from PE to CE but packet loss from PE to Internal network.It doesnt appear in any feeds, and anyone with a direct link to it will see a message like this one.Upvoted Log in or sign up to leave a comment log in sign up Sort by best level 1 14 points 14 days ago So, in short, has anyone ever lost packets on their network I think youre going to have to be a little more specific about what youre seeing and what youve tried so far.
We had an interesting case where we were streaming about 20mbs video via multicast. The server was at 1Gb and the network was at 1Gb until the access ports which were at 100Mb. The port buffers would fill after a minute or so and start dropping causing the video to become choppy. After diagnosis, we set the server to 100Mb and the issue went away. Its not always about throughput, sometimes the switch has to introduce space between the packets and it uses a limited buffer to do this. You will never know until you identify the three things mentioned earlier. I suggest checking one hop at a time to find out where the loss is occurring. Try running a tracert x.x.x.x (on Windows) or a traceroute x.x.x.x (on Linux or Cisco or Apple) command substituting x.x.x.x with the IP address. Then check each hop along the way to see where the loss is happening. If one of the hops has a much higher response time, it could be the problem one. From each network device, type show ip route x.x.x.x (or show route x.x.x.x if using a Cisco ASA) for both the source IP of your traceroute and the destination IP where the packets were being lost. ![]() If you see drops then you probably had too much traffic going through that interface at once. Usually the WAN speeds are slower than the LAN speeds, so it is unlikely to be the cause, but it can be if the numbers keep going up when you press the up arrow and enter to make the command show again. Output errors will be increasing also if you are sending much too much traffic for the interface speed. This only matters if the CRC errors keep going up after pressing the up arrow again. If one side has half duplex and the other has full duplex, you will see runts on one side and collisions on the other. Make sure you dont have a side set statically to half-duplex. You may need to statically make them both match if it doesnt auto-detect and match. If none of those show a problem, you can try typing show log on each network device, then Google, or ask here about any error messages that show up. If its not, try isolating where the packet loss is occurring by sending a lot of pings to each hop along the traceroute to see which one is dropping, and give as an overview of what the network topology is so we can give some more suggestions. If no, then get a capture at the host and CE box simultaneously and check what the difference between pings from your CE and what the PE box is sending are.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |