Cloud Connection is unavailable?

It’s really strange that you can get to and resolve the HE cloud instance from the Pi but your HE hub can’t connect.

You did the traceroute on the Pi with the port, like in the example I provided?
It’s important to verify that it can hit that TCP port and can receive the return responses, if your hub can’t connect to it, the cloud traffic won’t flow.

Could be something with the hub, your network, or your ISP (doubtful but possible).

Does your hub show up when you go to portal.hubitat.com? Not talking about when you login using hub IP.

Do you still have a ticket open with support?

Unless you have a way to check that the traffic is at least flowing to and from the hub (wireshark or similar) it’s going to be hard to determine what is up since you have checked or done most things already. And you are not having other issues.

By chance does the router show traffic connection logs?

Easy questions first:

  1. My Hub doesn't show up in the portal.hubitat.com (even though it's there - at 192.168.1.53).
  2. traceroute from the RPI (complete):
  3. Yes, I have opened a ticket with support.
  4. My router (Asus makes amazing routers, imho) does have traffic connection logs, but I haven't gone through them. I will start doing that now.

Is IPv6 enabled on your router?

maybe time to try something a little different...

You have a rPi and I'm going to guess you're using it wireless...

I'm thinking you can set the rPi to use the Hubitat address and move the ethernet wire from Hubitat to your wired port on the rPi.

That would give you the ability to test exactly what's going on with your network on that specific address.

A 'regular' rPi will switch from WiFi to Wired when you plug in the wire. But it will get it's own DHCP address. Two choices: alter your router to give Hubitat's address (192.168.1.53) to the rPi for this test; or, setup a static. It would be ideal to test via changing your dhcp reservation.

For setting up a static:
look at /etc/dhcpcd.conf and you should see the sample:
# Example static IP configuration:
#interface eth0
#static ip_address=192.168.0.10/24
#static ip6_address=fd51:42f8:caae:d92e::ff/64
#static routers=192.168.0.1
#static domain_name_servers=192.168.0.1 8.8.8.8 fd51:42f8:caae:d92e::1

You don't need the ip6 portion. So you'd want to alter just:
interface eth0
static ip_address=192.168.1.53/24
static routers=192.168.1.1
static domain_name_servers=192.168.1.1 8.8.8.8

Well, that seems like an interesting suggestion. Let me tell you how I did it:

  1. I don't have IPV6 enabled. (I'm just not ready to convert everything over.Yet.)
  2. Since I reserved a dhcp for my pi as I did for my HE, I just updated the table, and reserved the spot (.53) for the RPI instead of the HE.
  3. I gave the RPI an ethernet connection, and rebooted the router.
  4. I puttyed into the RPI and ran the following traceroute (I assume that it's to HE cloud headquarters):

This shows that my RPI has no issues getting to the HE cloud, but my HE cannot.
Is that a correct analysis?

You unplugged the HE to not have duplicate address, right?

And I did two Curl's to test it further:

$ curl 13.59.2.98:8883
curl: (52) Empty reply from server

$ curl https://13.59.2.98:8883
curl: (35) error:14094412:SSL routines:ssl3_read_bytes:sslv3 alert bad certificate

In other words, you probably should see the same. Errors BUT they are errors from the other end.

It does seem at this point to be an issue on the hub side then.

Try using curl like @csteele did.

Did the traffic (past) logs in the router show anything while HE was still plugged in? Your should see the hub connecting out on a high port to the HE cloud instance on port 8883 and a response back. Note: the IP might not be the same as the one I gave you.

As long as it is made by Corning, I'm happy! (this is my 28th year making optical fiber at Corning! :wink: )

3 Likes

Nice! I had a chance to take a fiber course, but didn’t, wish I did though just to learn more.

1 Like

E) All of the above. :wink:

I would always take MM over SM. If I remember correctly, SC and LC are simply connector types like MTP and FC, right? (It's been a few years since my datacom days, so I'm a bit rusty on my fiber technologies).

It does appear that the issue is at the HE cloud side.
Since I have been able to get to the HE cloud.
It appears that I'm just not "let in".

I guess support is the only ones that can troubleshoot is further, maybe reset your registration?

Support has come through!

They have figured out the problem.
In order to start this hub (a new C-5), I downloaded my C-4, and uploaded it to my new C-5.
All devices (of course) have to be excluded and then included to the new hub.

However, my original C-4 cloud keys are on my C-5 Hub!

They are currently trying to figure out what to do about that.
Mystery has been solved!

5 Likes