Hue bridge logging / troubleshooting

Is there any logging facilities or way to troubleshoot the hue bridge (built in) integration?

I only have 3 bulbs on the bridge and some lighting automations. It all works 99% of the time, but occasionally a bulb won't turn on from an automation. I can see in the events for the bulb that the automation sent whatever command. In this example "Command called: setColorTemperature".

Normally I'll see the response from the bridge a second or two later. In these cases it doesn't work, I get nothing else in the events.

I don't think debug logging will help since it's random how often it happens.

Based on the details you shared, the command leaves the Hubitat hub but gets lost within the network. This is not common but have seen few reports. Do you have any other network connected devices? Can you control the respective bulb from Hue immediately after the said failure? Is the bulb listed as not reachable in the Hue app? And last but not least, are these Hue bulbs or a different brand connected to Hue bridge?

They are actual hue bulbs.

I did open the hue app up after one failure and the app showed the bulb in the same state it was actually in.

Even if I cause Hubitat to control the bulb again seconds later, it turns on/changes just fine. Manually toggling it from the device page is just about as responsive as accessing it directly from the hue app.

I've never seen any of the bulbs show un-reachable.

Hue hub and hubitat are in the same network, on the same switch.

I think the only other network devices I use with Hubitat is thru the Kasa integration. Once in a great while I'll get a missed automation there too but it's extremely rare.

1 Like

If the Hue hub seems to be working fine when using the Hue app but commands seem to be dropping on their way to the Hue, I'd definitely look at network connectivity. As @bobbyD points out this isn't that common but it's also not unheard of.

If you have any network diagnostic tools you might try using them to test basic connectivity from your computer to HE and from your computer to Hue. Pingplotter is a simple one that works pretty well for me. If you don't or aren't comfortable using them, you can ping both HE and your Hue hub from your workstation for a while and see if you see anything odd. Any dropped packets would definitely be an indication of an issue. It is possible to monitor traffic directly between HE and Hue but it requires relatively sophisticated network knowledge, tools, and a managed switch.

What kind of switch is it? Is it managed or unmanaged?

@bobbyd out of curiosity does the native HE integration with Hue use HTTP like CoCoHue? (UDP for discovery.) I'm assuming it does since that appears to be the only connection method if I'm googling correctly. But you guys are clever so you never know.

I just ran about 5 minutes of a constant ping to both the hubitat and hue bridge. All looks good at 1ms or less and no loss.

I do have a managed d link switch. Just for giggles I looked at the port that hubitat is connected to and saw a bunch of TX and RX errors since who knows when the switch was last rebooted. I cleared the counters and after an hour had a couple hundred errors again on that port. Swapped the patch cord (was using the one that came with the hub) with a better premade cord and reset the counters. A bit later I checked back and I'm up over 1000 tx errors again. I'm swapping it to another port right now to see.

FWIW, it's a gigabit switch. The hubitat switch port is set to auto/auto, but is currently linked up at 100m/half duplex. Hubitat is set for "Force 100" with no mention of duplex.

Rather interesting. I've never looked at those port/speed/duplex/errors before. Never really saw much reason to.

On a side note, you've got me wondering if the built in integration is using UDP vs TCP. Could explain lost communications.

I am pretty sure it’s HTTP which means TCP but that doesn’t really mean much if it retries and eventually fails. However if you’re seeing errors in the switch that’s definitely an issue.

I just checked mine. HE is set to 100mbps fixed and the switch port reports 100mbps full duplex. What if you switch HE to fixed 100? If that doesn't work maybe you have a bad port somewhere. The issue is - switch or HE?

So with it set to forced, on the new port is was still linking up at 100 half duplex and counting up errors over an hour or so time.

I changed hubitat to auto, and now it's linked up at 100 full duplex. I cleared the errors right after that. About an hour later I have 22 rx errors. That's better than it was, but still alarming. Last thing I'm rebooting the hub, the switch, and letting it soak for awhile.

Yeah that is better but there's still something not quite right. I mean I suppose it could be a hardware issue on the HE side but ...

I just checked the switch port my primary HE is connected to, zero errors, counters go back a month. On some of my other HEs I do see a couple of errors but <=10 in the past month, which doesn't seem to be an issue. In my other life I have seen endpoints drop packets for reasons other than network... like memory issues etc. But I have not heard of such a thing on HE. Will be interesting to see what your reboot shows.

Are you on the latest D-Link firmware? Do you by any chance have jumbo frames enabled? I know that causes problems for HE but they usually exhibit themselves by locking up the hub. Also if you have DPI enabled (not sure if the D-link supports it) you might try turning that off.

So on the new port that hubtat is connected to has 295 rx errors since last night. I actually turned off jumbo frames last night as well because they were on.

Not sure what DPI is, but I don't have that option. Just Speed/Duplex/MDIX (crossover)/and jumbo frames. Pretty run of the mill stuff.

The errors definitely follow hubitat. But I'm not really convinced if that's 100% the problem or not.

Just for giggles I knocked that port down to 10 mb to see if the errors continue.

[EDIT] I was just poking thru the system log in my switch. The hubitat port is flapping sometimes several times an hour.

Wonder if there's anything support can look into with that.

Reach out to @support_team to see if they are able to assist.

1 Like

As @ogiewon suggested, please send us a PM with your MAC address and we could check the engineering logs to see if something jumps out.

Update. Just saw you sent the PM as I was hitting reply :slight_smile:

1 Like

AH yeah that's not helping.

DPI = deep packet inspection.

Gotcha. There are some broadcast storm stuff and DDoS protection but all that stuff is disabled.

Doesn't appear to be flapping set to 10/half but the errors are counting faster. I have a line of communications open with bobbyD now.

1 Like

Not to hijack the thread, but did you ever get a chance to see if you could see anything on mine? I sent ya a PM with my ip/mac.

I was the one seeing errors on the switch interface facing hubitat.

user3768,

What makes you suspect the ecobee app? I'm running it without any real issue. At least not that I know of. I know sometimes it's up there in resource usage though.

Nothing too unusual, although I have one item I had to run by our engineers. Will get back to you on the PM once I have more details.

2 Likes

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.