Hi guys, I have a question about lagging of ST app. After i have installed a few additionals power plugs with metering option, the system became unstable or i can say unresponsive. For instance, when i hit power on button withing ST android app, it often takes 2 minutes for that device to turn on. I have also two integrations with Sharptools and constantgraph.
I can’t figure it out, where to start looking. I don’t believe that those power plugs would do this…anyone can advice ?
IF i look into ST CLI log, there is a lot of communication, but it should handle this as far as I know.
I also have been adding a lot of power metering plugs and virtual power meters over the last 3 weeks. No issues until last night when I added constant graph. This morning all my Zigbee devices are going in and out and I can’t adjust temps on my Zigbee thermostats. I have 3 different hubs and all my Zigbee devices are on one and the virtual power meters on another. This has been working without issue for the last couple of weeks. I just went and removed my token from constant graph and I unplugged all of my hubs and I’m waiting for 30 minutes to plug them back in. Should rebuild the Zigbee network. I know I only had to unplug the Zigbee hub but I just did them all. I read a post from JD where he said polling from constant graph could cause issues if it was polling to often. I really don’t know how to set it up, so I may have done it wrong.
I guess it was not a good idea to unplug the hubs for 30 minutes. Now all my Zigbee devices are going crazy and show online and then offline constantly. it settles down for 10 minutes or so and then starts again. Not sure what to do now. Maybe one of the Iris plugs I added recently with Zigbee switch power MC is acting up and causing Zigbee mesh problems, but its been fine for a few weeks. This just started this morning.
Did you unplug all of your zigbee devices? That’s not how you do a zigbee heal. Instead, you take just the hub off power (unplug and remove any batteries if you have them), but leave all of the other devices on power. That way, the other devices will go into “panic mode” because they can’t find hub, and when the hub does come back on power, they will start rebuilding their neighbor tables. This can take a few hours, so you may not see results until the next day.
If instead, you take everything off power, you create a different situation. The end devices never go into panic mode, because they never had a time when they couldn’t reach the hub. Instead, when you put everything back on line, it just starts using the old routes, but everybody has a lot of catching up to do and, yes, you will see really weird results for a little while. Nothing you can do except just wait it out.
Once everything is fairly stable again, if you do want to do a Zigbee heal, take just one hub off power at a time, leaving everything else on power. That creates a very specific condition related only to routing when the hub goes back on power. Again, though, it can take about a day before that whole process completes. But at the end of it, your Zigbee mesh should be at optimal routing for your setup. I can’t speak to any other smartthings architecture issues, of course.
JD, thanks. I only removed power from my 3 hubs, not any of the devices. I know I should have only done the Zigbee hub but I figured why not lol. So now I removed the 8 Iris Zigbee plugs that I added to monitor power in the last couple of weeks and now everything is stable. I never really played with the power monitoring much and didn’t know that adding to many would cause issues like Paul mentioned. I thought that being I had separated all the Zigbee devices on its own hub that there would be enough resources for the power monitoring. I guess I’m going to have to go a different route to accomplish this. The thing I don’t understand is that it worked fine for the last couple of weeks and then all the sudden this morning things got bad.
Mesh networks are intended for tiny infrequent messages, like a light switch going on or off. It’s way too easy with power monitoring to flood the network and cause a lot of issues. In fact, one of the easiest ways to take down a Z wave network is to put two Aeotec 5 monitoring smart plugs in the same room. Lots of forum reports on that one, they would flood the network and nothing could get through.
Anyway, yeah, Wi-Fi or any star topology architecture is likely to be much better at handling these.