I woke up this morning to a set of lights normally controlled by Simple Automation that should have been on, but were not on. Any ideas on how to troubleshoot this?
The automation has been reliable for the last couple of weeks. (It's a holiday light automation, so it's fairly new.)
The devices (2) are reliable as well. One is directly in sight and routed directly to the hub. The other device is not a direct connection, but its route is also reliable.
The automation is on a C5 hub, while the devices are on a C7 hub connected/shared via Hub Mesh. Both hubs are running .156.
Two other Simple Automations ran fine this morning, both before and after this particular one was scheduled.
The automation fired on the C5 hub, but the devices on the C7 hub did not respond. (Or, at least there was no logging of any issues.)
I manually (but electronically via each individual device page) turned the devices on ~40 minutes after the automation time and they turned on very quickly with no issues.
I'll post some screenshots below, but I couldn't find anything in the logs that pointed toward any particular failure. As I mentioned, the automation logged that it fired, but there's nothing in the device logs at that time (6:00AM,) either positive or negative.
Help! Please... This is a bit troubling because I want to logically move another receptacle (physically, add a new 700 device and retire a 500 device) over to the C7 (leaving the automation itself on the C5,) but it's more critical that this future automation operate reliably. (On that note, it gives me a bit of an impetuous to do some automated status checking at various times during the day.)
For what it's worth, I just had a Simple Automation rule fail to fire. Just like you, it has been extremely reliable. All it does is turn on an outside light at sunset. Looking at the event log for the light - nothing. So, I don't think the rule fired.
Logging was not turned on, so I turned that on. Now, I'll wait and see what happens at sunrise, when this rule should fire again to turn off the light. I'm not sure what else can be done to troubleshoot.
A symptom for hub mesh failure would be a device event appearing on destination hub (where linked device is) and same event not appearing on source hub (that physical device is connected to). You'll need to turn on device logging the source hub, that will enable logging on destination hub, too. That can be double checked in device details.
Edit: I don't see any hub mesh errors in the logs for either hub.
Not updating to .158 wasn't a conscious decision, I just haven't gotten there yet.
Obviously, it's hard to disagree with mesh issues, but both these specific devices and the mesh as a whole have been fine both before and after this event. Those specific devices were fine 40 minutes later and another device on a different Simple Automation were fine ~1:20 later. /shrug I was hoping there was a hidden smoking gun in the logs somewhere.
I thought I had everything logging-related enabled for normal use, but I'm wondering if you saw (or didn't see) something in my logs. Thanks for the follow-up!
Grrrr... Something's up. Same thing last night, only with a different simple automation controlling a different device. (But same scenario: automation on the C5, device on C7, automation fired on the C5, but the C7 had nothing in the logs.) ETA: And both hubs are now on .158.
ETA: I figured it out. Still an issue, so I'll leave the post intact, but a different issue. I was messing around with driver assignment yesterday in order to get the debug logging to 'stick' and not time-out. One of my devices (and, of course, only one device, but the one that was tied to the automation above,) missed the driver change from the 'bad' driver back to the 'good' driver, and the automation didn't have anything to control. Figured I'd leave this in case gopher.ny wanted to look at that issue. It's not a huge deal, I just need to double-check that the new driver gets communicated back, should I perform that exercise again.)
Correct. The device page on the source hub had the correct driver listed (and all of the correct device commands,) but the destination hub still had the previous driver listed.
I don't think I did, simply because it had been ~10 hours since I finished those changes, and as I said above, the source hub had the correct driver on the device summary page (and within the device edit page, too, plus all the correct device commands listed.)
Anything's possible, though. My CRS can be strong at times.
This is a bit troubling because I want to logically move another receptacle (physically, add a new 700 device and retire a 500 device) over to the C7 (leaving the automation itself on the C5,) but it's more critical that this future automation operate reliably. (On that note, it gives me a bit of an impetuous to do some automated status checking at various times during the day.)
