Random Problems with small section of C5's is likely a QC issue at factory

WOA!
Let me get this straight. Your contention is that Hub slowdowns are caused by CRC errors in the NIC of the HE. And , you were able to see these errors because you have a managed switch to which the HE is plugged into. The managed switch has some sort of report which you copied to above (I saw a few of your screen shots).
Is that correct?
I've checked the Asus and Synology routers that I have, and none of them show any CRC errors. What kind of switch was that?

Yes and no...

I believe the problems with my specific hub, and maybe with a small subset of other hubs encountering both auto negotiation and CRC errors is due to a QC issue at the factory. (it happens all the time, nothing HE is responsible for and something that would be very hard to catch if its only affecting a handful of units)

The generic slowdowns some have after a few days I don't believe have any relation to this, and are likely something to do with a bug in driver or app code somewhere, or with devices sending ill formatted data, or broadcast storms on the network.

I actually just assumed the hubs were a bit on the slow side UI wise and could benefit from stronger hardware until I got to see what a hub without those issues can do.

Routers usually don't report CRC errors, even DD-wrt doesnt report them anywhere in the interface. Managed switches are your friend here, not sure if there is a software package for detecting them. In most cases CRC errors are down to bad hardware, usually cables of a bad NIC.

1 Like

That's a reasonable interpretation of your data.

Yup.

Well the hub will be on its way back to bobby and the rest of the team to inspect as soon as the replacement arrives and i transfer my setup over. With luck they will be able to confirm where the problem came from and prevent others from encountering it in the future, or at least have an idea of a potential cause of similar problems in the future

One positive in this for me is in trying to resolve the issues I finally got around to updating my entire home network. New routers, switches, and Solid core (not CCA) cat 6 cabling.

3 Likes

Update

New hub arrived yesterday. No CRC errors, instant 100F connection to my switch. Everything is so much faster!

Lights come on pretty much instantly after the sensors fire, and so far no slow downs.

This is the experience I was expecting based on everything I had read... Now I can really start to see what this little thing can enable me to do :slight_smile:

Thank you @bobbyD and @chuck.schwer for taking care of the issue so fast once you were actually aware of it! You can now consider me a happy customer looking forward to seeing what you guys come up with for us next :slight_smile:

3 Likes

So the question now is - was this a random faulty hub (perfectly acceptable) .. or have more people got this same issue?

2 Likes

and how to detect it if you dont have a fancy switch to show you?

Hi
I have the C5 hub connected to a smart Netgear switch. I have done the "fix" for the Netgear switches. My hub it connected to port G13. A can be seen on the below image it is the only port with packets received with errors as well as Collision Frames. I have tried another network cable as well as another port on the switch, same result.

Are you having any symptoms, slowdowns/lock ups?

I don't have many devices (+_ 15) so I don't think the hub is being "stressed" but it does worry me with why there are any errors at all. I will be adding devices over time and don't want this to be an issue. I have had an error 500 about 3 times.

Not sure if this means anything to anybody, this is the detailed stats for that specific port

Might be worth tagging @bobbyD to have a quick look and advise

Sorry not sure how to add tag after the post? :roll_eyes:

any one know if this swithc would detect these errors?

That should as it is a smart switch, which is similar to my Netgear Smart switch.

couple quick questions for you. Are you using flat or round cables? patched or non patched? and solid core or stranded cat cables?

Flat cables=more prone to interference
non patched can cause issues if the switch isn't perfect at switching the rx/tx pairs for you
and solid core cables require special connectors for a stable data stream... a lot of manufacturers use 2 pronged rj45s, and you need the 3 prong with solid core as 2 prong connectors can miss or break the wires, where the 3 prong actually hug the wire giving a stable signal.

1 Like

Are you using any wifi devices? are your switch or router providing access to a hard drive? is your router near a microwave, water heater, electrical box, water pump or heating/cooling system?

I grabbed a 200ft extension cord and took my hub router and switch as far away from the house and other devices i could to ensure it wasn't an issue being caused by anything in my home.

Eliminate every outside cause you can, it will help support as they won't have to go through all that troubleshooting/questions, and will help speed up getting a replacement if it is determined to be the hub :slight_smile:

Hi digitydogs

I am using cat6 round cable, non patched, directly in to the switch, I have tried a cat5e patch cable as well same result. I find it strange that I have about 12 devices plugged into the switch and only this device shows the errors, not 1 error on any of the other ports, which are all using the same cables in to the same switch, have tried different ports as well. I even ran a temp cable directly across the floor and had the same results.

have plenty of wifi devices, running Ubiquiti unifi router and AP's, as well as Mikrotik. I have a split wifi network between 2.4Ghz and 5Ghz. Hubitat not near any microwave or heating or cooling.

good sounds like you have covered all the bases then. that will save you some trouble in getting this pinned down. As much as I hate to say it I hope @bobbyD sends you a new hub and has you send that one back. If its the same issue then maybe between testing our hubs they will be able to pinpoint the cause of some of these random issues once and for all and get the platform rock solid!

However, could this be the source of the mysterious slowdown virus that has affected many?
Is there ANY way to diagnose this error besides a fancy managed switch?

Is there anyone out there, with significant network analysis skills, that can suggest a way to test out this theory? (That some C5 hubs have CRC errors due to a quality control issue.)
P.S. Just by coincidence, I just got a fancy TP-Link 16 port switch. Before I open it, I'm going to check the manual to see if it will report that stuff.