Jump to content

Welcome to Smart Home Forum by FIBARO

Dear Guest,

 

as you can notice parts of Smart Home Forum by FIBARO is not available for you. You have to register in order to view all content and post in our community. Don't worry! Registration is a simple free process that requires minimal information for you to sign up. Become a part of of Smart Home Forum by FIBARO by creating an account.

 

As a member you can:

  •     Start new topics and reply to others
  •     Follow topics and users to get email updates
  •     Get your own profile page and make new friends
  •     Send personal messages
  •     ... and learn a lot about our system!

 

Regards,

Smart Home Forum by FIBARO Team


  • 0

Odd network issue


Cameleon

Question

Hi everyone,

 

I have an odd issue since some recent network changes and it appears to be down to my HC2 - or something it communicates with or communicates with it. 

 

So long time working setup I've had (UK based) is broadband internet from PlusNet (BT) and a BT Smart Hub 2 router

 

Broadband router <<eth>> Netgear GS108 switch <<eth>> [wired devices]

 

where [wired devices] are HC2, rPi/homebridge, Philips HUE, Windhager (pellet boiler), Heamiser (UF heating), Synology NAS. 

 

Recently we've needed more speed as although we're on FTTC, we only get 15Mpbs download and 1Mbps upload. So we've moved to a LTE router in the Huawei B818. This thing flies compared to the BT connection, 130Mbps down and 30Mbps up. Initially for testing, as it is sited the other side of the house for the external antenna,  I had it set up on an old Vodafone router as an AP (disabled DHCP and all that) and that only had internet access, all the other setup as above remained and had to access them on 2 different AP/networks. 

 

Now I've tried to merge them but something isn't happy. Connecting the B818 to the GS108, we get some really weird behaviour. Internal LAN connections are all fine and super speedy. WAN based connection (i.e. the internet) are really unstable to the point that I have to refresh web pages multiple times to get them to load and sometimes they just dont load at all. Speedtest.net app sometimes connects fine and gets full 130/30Mpbs and othertimes fails completely, or download is fine but upload fails. 

 

Through a process of elimination, I have worked out it is when the HC2 is on the 'new' network that the connection goes weird. But I have no idea why it would do this?! IP is set to manual, points to the new router properly and the internet light shows on it. Changing the IP or swapping to DHCP works for a while then it freaks out again. Again, all internal connections are fine, no issue getting to the management page on the device. 

 

I dont have another dedicated switch to test with, but as a quick test I have hooked up my old BT SH2 and running the B818 to it and then onto the HC2, it appears to be working (as I type!). It only has 4 ports so I can only test another 2 devices on there with it, so can do some limited testing. 

 

Any ideas? And how to go about nailing down the cause? Dodgy switch? Dodgy HC2?

Link to comment
Share on other sites

16 answers to this question

Recommended Posts

  • 0

This is a long shot but can I tell you my gut feeling (as an ex network engineer)? The symptoms scream "duplicate IP", possibly two gateways claiming the same IP. This will often lead to a internet connection that flips between "nothing working at all" and "everything fast and OK". like "ping 8.8.8.8" going from nothing to "a few ms" with nothing in between... There is nothing special about the ethernet of a HC2 and I haven't seen anyone post anything remotely similar to your symptoms. But please do check the cable as well, you'll need a CAT5E - probably not an issue because it is hard these days to find patch cables that are worse than CAT5E.

Link to comment
Share on other sites

  • 0

It feels like "duplicate IP", as @petergebruers said. Broken cable or bad nic chip (HC2 is usig intel nic, not known for any problems) will not influence switch operation. On that layer only PoE might be issue (when you connect HC2 to PoE port - if your GS108 is PoE version).

Link to comment
Share on other sites

  • 0
  • Inquirer
  • Hi guys,

     

    Thanks for the replies. Happy to take on any investigation if it helps resolve the issue. It really is very weird. I'm by no means a network expert so could well be something misconfigured somewhere and when running the old broadband router in the mix, I would have thought it even more likely as you never know what they are running under the skin even when you tell it to turn everything off and just be pass-through. 

     

    GS108 is the standard non-PoE version.

     

    I don't think I have any duplicate IP's. Looking on the B818 device list, I assume everything should show there? The devices are a mix of static IP (for HA stuff mainly) and DHCP for those I can't/havent configured for static and wifi devices (laptop, phone, etc). 

     

    B818 -192.168.8.1 (Hauwei default)  - router/DHCP/DNS

    BT SH2 - 192.168.8.254 (using as switch, everything disabled)

    HC2 - 192.168.8.11 (static) - has .1 as it's gateway and DNS

    NAS -  ‏192.168.8.2 (static) - has .1 as it's gateway and DNS

    Hue - 192.168.8.5 (static) - has .1 as it's gateway

    Neo -  DHCP (Headmiser) - can't change to static - no option

    RPi - DHCP (Homebridge) - could change to static if required

     

    So I can't see any issues there. There are no duplicates and gateways/DNS seem to be correct. 

     

    Could it be the switch that is bad? Not sure how to test that? To rule out the GS108, I've hooked up the B818 and HC2 to the BT SH2 in AP mode, to just use it switch functionality. Seems to work. To then check the other details, I've daisy chained the BT SH2 to the GS108 with the remaining other wired connections (NAS, Hue, Neo, RPi) and it seems *touch wood* to be stable. I'll leave it like that for a few hours and see. Usually it bombs out quickly. When everything is on the GS108 as soon as I boot the HC2 the connection dies. But nothing is on the same IP as the HC2 and I've even tried changing the HC2 IP and it does the same - both on static and with a DHCP reservation on the B818. 

     

    ?

     

     

    Link to comment
    Share on other sites

    • 0
    16 minutes ago, Cameleon said:

    Looking on the B818 device list, I assume everything should show there?

    No, that is the devious thing about duplicate IPs, they do not show up on a list, you might be able to spot it by looking at ARP cache. I am on a mac so here's a MAC example but iirc the same command exists on windows (any version).

     

    arp -a
    ? (192.168.0.1) at 90:5c:44:ae:aa:aa on en0 ifscope [ethernet]

     

    So this tells me, ethernet port with mac address 90:5C...AA is the one my mac thinks is IP ....1 and my gateway... This is correct (I can verify that by looking at the router label)

     

    If you have accidentally assigned 192.168.0.1 to another ethernet card, random machines can fail to ping internet machines like 8.8.8.8

     

    If you have such a problematic PC, arp cache might show you:

     

    ? (192.168.0.1) at 48:f:cf:a7:34:75 on en0 ifscope [ethernet]

     

    The MAC is not the correct on. Now you know for sure .48:...:75 has the address of the gateway

     

    You can do "arp -d" to flush that cache and ping to get a new pair of IP/MAC...

     

    16 minutes ago, Cameleon said:

    Could it be the switch that is bad?

    Never say never but I give that a 1% chance. It is an unmanaged switch, so very few diagnostic tools.

    If a cable or connector is bad , ports can flip between "on" and "off". But that should be limited to a single port.

    I have a managed switch, and if it sees too many errors, it disables the port to avoid this kind of "flapping"

     

    16 minutes ago, Cameleon said:

    Usually it bombs out quickly.

    Yeah duplicate IP conflicts depend on the speed of the machines answering the ARP request, the fastest machine wins. So there is some randomness to it and also depends on routers our switches. Hard to confirm or rule out at this point, you've changed the topology.

     

    Connectors of ethernet cables can go bad... Or have a bad connection then start to work again when you bend the cable in a certain direction.

     

    I had a case of "oxidised contacts" myself... Difficult to diagnose and caused random 20 second delays.
     

     

     

     

    Edited by petergebruers
    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • Thanks for the feedback. The switch is an unmanaged (and fairly cheap and cheerful I guess at only ~£30) but it is meant to be one of their lifetime warranty ones so if it is bad then I can get it swapped out. But being unmanaged it will be hard to prove whats wrong if it is the switch. 

     

    One thing to note, I did have the old BT SH2 router set to be IP 192.168.8.1 some time back to allow me to switch from the router/AP to the B818 for testing. Clearly that would have been causing an issue at the time as some devices would have had the wrong MAC against the IP? But now that is set to 192.168.8.254. So I wonder if there are devices that have't been restarted when that change over happened and are 'hold on' to the old/incorrect MAC address?

     

    arp -a (on my Mac)

     

    ? (192.168.8.1) at e4:26:8b:ba:a3:59 on en0 ifscope [ethernet]

    ? (192.168.8.5) at 0:17:88:23:8b:93 on en0 ifscope [ethernet]

    ? (192.168.8.11) at ac:17:2:31:d:db on en0 ifscope [ethernet]

    ? (192.168.8.103) at b8:27:eb:30:f:7e on en0 ifscope [ethernet]

    ? (192.168.8.109) at f8:27:93:ef:1b:48 on en0 ifscope [ethernet]

    ? (192.168.8.127) at fe:7c:a2:4:51:2b on en0 ifscope [ethernet]

    ? (192.168.8.128) at ae:ff:14:a6:3a:21 on en0 ifscope [ethernet]

    ? (192.168.8.132) at 70:ee:50:1f:65:24 on en0 ifscope [ethernet]

    ? (192.168.8.133) at f8:e9:4e:a3:32:fd on en0 ifscope [ethernet]

    ? (192.168.8.138) at 70:ee:50:1f:65:24 on en0 ifscope [ethernet]

    ? (192.168.8.141) at ae:ff:14:a6:3a:21 on en0 ifscope [ethernet]

    ? (192.168.8.143) at d8:80:39:ae:3c:ff on en0 ifscope [ethernet]

    ? (192.168.8.145) at f8:e9:4e:a3:32:fd on en0 ifscope [ethernet]

    ? (192.168.8.254) at 78:dd:12:2e:4b:cb on en0 ifscope [ethernet]

    ? (192.168.8.255) at ff:ff:ff:ff:ff:ff on en0 ifscope [ethernet]

    ? (224.0.0.251) at 1:0:5e:0:0:fb on en0 ifscope permanent [ethernet]

    ? (239.255.255.250) at 1:0:5e:7f:ff:fa on en0 ifscope permanent [ethernet]

     

    I can see a few duplicates of mac in there where I've had devices such as my iPad on both the B818 wifi and also the U6 AP.

     

     

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • I can also see that on my Windows laptop and the HC2 (via the API) that the MAC for the router are all matching (e4:26:8b:ba:a3:59)

    Link to comment
    Share on other sites

    • 0
    4 minutes ago, Cameleon said:

    Clearly that would have been causing an issue at the time as some devices would have had the wrong MAC against the IP?

    Yes. And "timers" govern how long they keep the wrong address in cache...

     

    try:

     

    arp -anx

     

    This will show extra columns, two of them are called "Expire" and show how long your PC "sticks" with this IP/MAC  pair.

     

    8 minutes ago, Cameleon said:

    So I wonder if there are devices that have't been restarted when that change over happened and are 'hold on' to the old/incorrect MAC address?

    Yes. But arp timers are set to some "reasonable" timeout. IDK from the top of my head but probably anything from 10 to 60 minutes is common.

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • Thanks Peter. The ‘-axn’ doesn’t work on Windows but the output on macOS made sense. Seems they are limited to a few minutes before clearing. 

     

    Something suss with the HC2 though. I have hooked up the BT SH2 as a switch (so no WiFi, no DHCP, no firewall) and having that between the B818 and U6 (for WiFi) and it all works fine. So B818 <eth> BT SH2 <eth> U6 <wlan> Clients works fine. Rock solid, pages load quick, Teams calls spot on. Add in the HC2 to that on another port of the BT SH2 and seems good to start, but then Speedtest.net app starts to not get provider, or download fails or download is fine and upload fails. Teams call also freaks out and disconnects. 

     

    I need to test it with other devices but this is what happened on the GS108, so I’m thinking its not the GS108. Not saying the HC2 is to blame per-se but something it is doing appears to be freaking out the B818. When it happens, even going on the B818 own WiFi (bit of a stretch so connection is not fast), that appears to have the lag too. Need to try and get another Ethernet cable to the B818 and see what that’s like when it happens. 

     

    The other thing I’ve seen it freak out with when the HC2 is connected is if I got to the iOS App Store and check for app updates. This appears fine when HC2 is not connected but with it connected it seems to break the connection and I have to restart the router to get it all back to full speed. Saying that, I’ve just done the App Store refresh and its currently fine! So who knows!

     

    Could also be unrelated but changed the IP from static to DHCP back to static and the static setting wouldn’t stick. Had to do it a few times in a row for it to accept the IP address and not just go on DHCP still.

     

    Any ideas on what else to look at? Wonder if the B818 is just not good as a router? I mean I don’t have a lot of devices at all, but it seems to be solid when you just have the U6 clients on it or Wifi from the BT SH2, but add in the HC2 and it seems to drag it down. Weird as I had no problem on the BT or Voda routers previously! 

     

     

    Edited by Cameleon
    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • So having been rock solid all afternoon with just B818, BT SH2 and U6, added back the HC2 and in less than an hour the internet is freaking out. iPad and iPhone can’t load we pages first go, Speedtest.net app fails, but internal stuff is all fine, can get to all the devices and speed is fine to them. So think HC2 is causing some issues on the WAN connection of the B818. If I reboot the B818 it is fine again for an amount of time. Without the HC2 connected (since 7am today) internet has been fine on the wireless via U6, and problems start less than an hour after plugging the HC2 back in (only the HC2).

     

    To get it all to behave again I have to disconnect the HC2 and reboot the B818

     

    Even trying to post this failed! DOH! 

    Edited by Cameleon
    Link to comment
    Share on other sites

    • 0

    hmm, BT SH2 is reported to have firmware issues, B818 should normally disturb z-wave and not vice versa (1m distane between both is enough). I remember some DLAN issues caused by HC2 power supply (fixed it with some extra 5m cable between, and for another customer with different power supply). But that's all, out of ideas.

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • I have no idea whats gone on with it. But now managed to convince myself I needed a Ubiquiti UDM - yes I was looking for an excuse - BUT - it appears to have sorted things. Put the B818 in bridge mode and hooked that up to the UDM. Everything else is still via the GS108 and then into the UDM but no issues so far. Internet actually seems to be faster, so I'm happy!

     

    So... B818 <eth> UDM <eth> GS108 <eth> all devices inc HC2

     

    Would have been nice to know what caused it not to play nice with the B818, but I think the HC2 was the issue. I think the GS108 is fine as I had all the other devices on it yesterday with the B818 (not via the BT SH2) and it was all working well until you added the HC2. Remove the HC2 and reboot the B818 and everything was fine again. It was very obvious adding the HC2 to the mix caused the network issues. And you had to reboot the B818 after removing it. All the same setup now but with the UDM between the GS108 and B818 and its all fine with the HC2 on the GS108.

     

    Very confusing!!!

    Edited by Cameleon
    Link to comment
    Share on other sites

    • 0

    oh well, execuse granted ^^

    Link to comment
    Share on other sites

    • 0
    2 hours ago, Cameleon said:

    Its been solid all night, running super quick so I am happy! 

    Thanks for reporting back. If you're happy, I am happy!

     

    16 hours ago, Cameleon said:

    It was very obvious adding the HC2 to the mix caused the network issues. And you had to reboot the B818 after removing it.

    That is reproducible? Then someone might be able to understand what is going on but I am not sure if it is worth the trouble and I am not saying that I would be able to help you. But for the sake of completeness... You could try this tool to capture the traffic on some ports of your network (but since you don't have a managed switch, you almost certainly lack a monitor port to watch the HC2 so that is a no-go) and have a look at broadcasts storms or anything unusual, probably at layer 2 but also possibly rogue DHCP offers (which might assign invalid addresses to devices which then lose connectivity). But you probably would have noticed wrong DHCP assigned IP addresses.

     

    Please login or register to see this link.

     

    A think that can cause havoc on a network is some kind of loop (eg 2 ethernet cables running in parallel between switches) in combination with (or without) "Spanning Tree Protocol" which can cause switches to partition the network. I have seen it happen on somecore HP switches (managed, in the year 2008, that is ancient history, LOL) due tor firmware bugs. But I cannot in any way, relate this to a Fibaro HC2. But as I said, it is worth mentioning for the sake of completeness

     

    Please login or register to see this link.

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • I'm not 100% its reproducible, it seemed to take variable amounts of time for the network to experience issues, but they did always seem to happen when the HC2 was re-introduced (either by itself as the only other network device or with all the other HA devices already on the network). I did wonder if it was a broadcast storm but the only way I had to tell (which again is not reliable) was watching the switch and it didn't appear all lights were blinking at the same time a lot. Definitely no loops in the network, its such a simple setup and happened even when I just had B818, VF router/AP, U6 AP and HC2, so was very simple kind of hub/spoke where everything radiated off the VF router. 

     

    The only thing I can see that may have made a material change in installing the UDM, is I change the IP address of the HC2 back to the 192.168.1.x subnet from 192.168.8.x range the B818 allocates (the B818 still being on default 192.168.8.1, UDM network now being 192.168.1.x). Could have reset something in the networking side of the HC2. Or the UDM is better at ignoring some spam that the HC2 is sending out. I havent got my head around all the features of the UDM yet so could be something to be seen there. 

     

    As this is our home/WFH network, I will struggle to move devices back onto the old setup for testing :(

    Edited by Cameleon
    Link to comment
    Share on other sites

    • 0

    Sorry, I don't have an answer, Cameleon, but I am wondering if the B818 has connection issues.

    I upgraded my LTE router (D-Link DWR-921) to a B818 and got the expected Broadband speed boost: 100-200 Mbps down and 30-60 up.  But also found it could not connect WhatsApp calls between some phones, could not connect to Kindle devices, and mucked up my SIP phone.  Is there something about it that is not quite right regarding connectivity?  I have had to go back to my D-Link which is not bad for a budget device but is showing the first signs of age in a cheap router (needing rebooting every so often)  

    Link to comment
    Share on other sites

    Join the conversation

    You can post now and register later. If you have an account, sign in now to post with your account.

    Guest
    Answer this question...

    ×   Pasted as rich text.   Paste as plain text instead

      Only 75 emoji are allowed.

    ×   Your link has been automatically embedded.   Display as a link instead

    ×   Your previous content has been restored.   Clear editor

    ×   You cannot paste images directly. Upload or insert images from URL.

    ×
    ×
    • Create New...