09-29-2020 08:03 AM
Hi, we are a manufacturing plant. Our manufacturing lines and other networks are not well segmented. We would like to better segment the assembly lines and also the office. We have dropouts reported on the assembly lines that impede production and make it hard to add new equipment without creating more problems. There is also a high level of dropouts/errors reported in our network monitor.
What are your rates, and can we setup a meeting with you?
Attached is a picture of some of the dropouts reported.
We are told that our switch logs show a lot of “flapping”
Below is a generic write up of some of the problems I made to ask help groups on the web.
“Solarwinds shows our core and the switch our hyperconverged system plugs into reports a lot of errors and discards (billions). We have a stack of 4 C3850s (our core layer 3) and a 3850 10GB switch. We have a hyper-converged (HPE OmniStack) system that is plugged into the 10GB switch. We don't have any routing setup. The port that was selected to be the blocked port for spanning-tree had been one selected since the system was installed 15 yrs ago. We have added about 12 switches, maybe more. Maybe we need routing and an new port to block spanning tree. I am not familiar with this level of network troubleshooting I am semi-new to networking, and have . I have a CCNA, but I don't think that is going to help me troubleshoot this. Unfortunately, this is a manufacturing facility with PLCs that do a lot of multicasting . The PLC manufacturing environment is not really separated from the regular network. It is poorly isolated by a VLAN. We have about 20 VLANS. I also believe our QOS settings could be improved. I believe the system has out grown itself. Is there a place I can begin to troubleshoot what we currently have?”
PS: Below are a couple more considerations, but I don’t think are big contributors to the problem.
“We do have a couple dozen cameras. We bought switches just for them. The plan was to plug just those switches directly into the camera controller and use a second NIC so we could view them. The problem is 3 of the 5 switches won’t fit in the IDF cabinets and we have to make room before we move them all. About half of the cameras are in the new switches, but they are not plugged into the camera controller yet. However, I just thought of 5 cameras that could be moved now in the IT room to the new camera switch in there. “
Solved! Go to Solution.
09-30-2020 09:47 AM
I will ask my boss because we have to have a non-discloser agreement. Two of the companies we work with said we need to start by upgrading the firmware from the 3.x to the 16.x family. They said this will eliminate many Xmit-err and multicast bugs. Thank you for your offer.
09-29-2020 09:13 AM
Hello,
post a schematic drawing of your topology, showing what is connected to what, the running configuration(s) of the problem switch(es), as well as the output of 'show interfaces x' of the interfaces that have a lot of discards.
09-30-2020 09:47 AM
I will ask my boss because we have to have a non-discloser agreement. Two of the companies we work with said we need to start by upgrading the firmware from the 3.x to the 16.x family. They said this will eliminate many Xmit-err and multicast bugs. Thank you for your offer.
10-01-2020 04:41 AM
"Two of the companies we work with said we need to start by upgrading the firmware from the 3.x to the 16.x family. They said this will eliminate many Xmit-err and multicast bugs."
Please post a follow-up, i.e. if upgrading the IOS versions does, indeed, fix your xmit and multicast issues.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide