Powerwall Connection Problems

[slazinger_7]

One of our gateways (on a two gateway system) seems to drop its data connection about every month or so. We do a reset of the gateway and everything will come back up just fine and run for a couple weeks, then it drops again.

Anyone else having a similar problem? And/or is there a fix?

The longer back story is that we set up our own remote monitoring using Gafana using a package at github. Periodically, the data will just stop reporting, i.e., the charts are blank. If we expand the history, we can see exactly when the data dropped. When this happens, we also cannot connect to the gateway at the wifi IP address, the ethernet IP address (we have both set up and active), and even connecting to the gateway directly will not work, i.e., we can connect to the gateway via wifi, but going to the IP address (192.168.91.1) just errs out.

The real-time data reporting that one can access using the Tesla app appears to work fine, but we can't make any changes to the Powerwall settings. For example, if we try to change the setting to charge from the grid in anticipation of an outage, the change won't take effect.

The other gateway, which is still connected to the internet via its own wifi, ethernet, and cellular connection works just fine (as expected), and never seems to drop - we can see continuous data through it's own Gafana monitoring, we can see the portal through its wifi and ethernet connections, and we can make changes to the settings just fine.

Seems there's something wrong with the gateway and its data reporting, but not sure if there's anything we can do to troubleshoot.

Any thoughts?

 

[jgleigh]

I would disconnect the WiFi and just use the Ethernet. My Gateway was very unstable until I hooked it up to Ethernet.

 

[zƬesla]

Is there a way to physically disconnect Wi-Fi on GW1?

 

[jgleigh]

Is there a way to physically disconnect Wi-Fi on GW1?

I don't know. I have a GW2 and I just logged out of the WiFi so it doesn't connect to my router anymore.

 

[dailo]

One of our gateways (on a two gateway system) seems to drop its data connection about every month or so. We do a reset of the gateway and everything will come back up just fine and run for a couple weeks, then it drops again.

Anyone else having a similar problem? And/or is there a fix?

The longer back story is that we set up our own remote monitoring using Gafana using a package at github. Periodically, the data will just stop reporting, i.e., the charts are blank. If we expand the history, we can see exactly when the data dropped. When this happens, we also cannot connect to the gateway at the wifi IP address, the ethernet IP address (we have both set up and active), and even connecting to the gateway directly will not work, i.e., we can connect to the gateway via wifi, but going to the IP address (192.168.91.1) just errs out.

The real-time data reporting that one can access using the Tesla app appears to work fine, but we can't make any changes to the Powerwall settings. For example, if we try to change the setting to charge from the grid in anticipation of an outage, the change won't take effect.

The other gateway, which is still connected to the internet via its own wifi, ethernet, and cellular connection works just fine (as expected), and never seems to drop - we can see continuous data through it's own Gafana monitoring, we can see the portal through its wifi and ethernet connections, and we can make changes to the settings just fine.

Seems there's something wrong with the gateway and its data reporting, but not sure if there's anything we can do to troubleshoot.

Any thoughts?

Are you connect to a 2.4Ghz or 5Ghz network? Or maybe a dual band one? Although the gateway supports 5Ghz I found the connection to be flaky and once I switched it to my 2.4Ghz only the connection has been very stable over the last year.

 

[BGbreeder]

Are you connect to a 2.4Ghz or 5Ghz network? Or maybe a dual band one? Although the gateway supports 5Ghz I found the connection to be flaky and once I switched it to my 2.4Ghz only the connection has been very stable over the last year.

AFAIK, the Gateway only supports 2.4GHz (see here), and there many reports of folks with 2.4/5GHz combined networks (single SSID) having issues. I suspect that the routers try to push the Gateway to 5GHz, which it can't actually do, causing it to hang.

@slazinger_7 I would start by making sure that your Gateway is on a 2.4GHz only network, or running an ethernet cable to it.

All the best,

BG

 

[slazinger_7]

I have it connected via Ethernet and wifi, and neither work when we have this problem.

Even if the wifi and/or Ethernet were dropping, it still doesn't explain why, when we connect directly to the gateway over its own wireless AP, the login page found at 192.168.91.1 still errs out as if it cannot be reached.

 

[dailo]

I have it connected via Ethernet and wifi, and neither work when we have this problem.

Even if the wifi and/or Ethernet were dropping, it still doesn't explain why, when we connect directly to the gateway over its own wireless AP, the login page found at 192.168.91.1 still errs out as if it cannot be reached.

Have you tried resetting it and see if that helps?

 

[jgleigh]

I have it connected via Ethernet and wifi, and neither work when we have this problem.

Even if the wifi and/or Ethernet were dropping, it still doesn't explain why, when we connect directly to the gateway over its own wireless AP, the login page found at 192.168.91.1 still errs out as if it cannot be reached.

My theory is the computer in the Gateway is actually crashing. The wireless AP module is still broadcasting, but the brains behind that aren't working. I get a full system shutdown in addition to no response to queries when my Gateway gets into this mode.

 

[slazinger_7]

Have you tried resetting it and see if that helps?

Yes - hitting the reset button inside the Gateway seems to restore all connections . . . for a while, then it all goes down again. This last time, the system was up about 24 days before we lost all connectivity again. As far as I can tell, the batteries are no longer going on during peak time at the moment. The installation is at a remote property, so someone is not always readily available to stick a screwdriver into the Gateway to do a hardwire reset.

My theory is the computer in the Gateway is actually crashing. The wireless AP module is still broadcasting, but the brains behind that aren't working. I get a full system shutdown in addition to no response to queries when my Gateway gets into this mode.

So it sounds like you have this problem as well? Any idea for a solution to fix? Another oddity is that one Gateway has problems, while the other doesn't seem to. Tesla is worthless for getting any tech support to fix anything.

 

[jgleigh]

So it sounds like you have this problem as well? Any idea for a solution to fix? Another oddity is that one Gateway has problems, while the other doesn't seem to. Tesla is worthless for getting any tech support to fix anything.

I switched to Ethernet and turned off the WiFi. I've only seen it crash once since then in over a year of stable operation.

 

[BGbreeder]

I switched to Ethernet and turned off the WiFi. I've only seen it crash once since then in over a year of stable operation.

If it has locked up even once on an Ethernet cable, I think that your Gateway has some Ethernet hardware issues. I would leave open the possibility that your router is the flakey party here, but I would have expected other devices to have issues.

Good luck,

BG

 

[slazinger_7]

I'll try switching off the WiFi so that the system runs exclusively off the Ethernet to see if that resolves the issue.

I don't think it's the router - we have a number of devices configured to run through the route, and they all work just fine (even our other gateway, which runs on both WiFi and Ethernet).

 

[BGbreeder]

Turning off the WiFi is a great troubleshooting step. In almost every Ethernet implementation that I know of, once wired ethernet powers on, the WiFi is idled, so, in theory with good cable, and connections, turning WiFi should solve your issues.

I guess if the cable, or cable connectors are making intermittent contact, the network subsystem could be switching back and forth between wired and WiFi leading to lockup, as the wired connection times out. Perhaps you could also check the sockets a cable ends, in case there is a piece of debris in there somewhere, or a wire contact is out of alignment?

All the best,

BG

 

[sovietaced]

I just ordered a Raspberry Pi so I will be pulling telemetry from the gateway myself but this behavior seems like it may be a memory/resource leak of some sort on the gateway side. The problem might be exacerbated by the telemetry polling as well as the engineers may not have anticipated the Gateway APIs being called so frequently.

 

[Matt-FL]

Something to check - Ethernet isn't always enabled on the Gateways, at least not according to my Tesla badged installer. I asked him to enable Ethernet, as I plan to hard wire it this fall (too darn hot now to climb in a Florida attic). You can see if it is enabled from the gateway web interface, under network:

 

[slazinger_7]

Disabling Wi-Fi seemed to solve the issue. So far no problems.