I only have a subset
Interesting I just forced a reboot and now its taking 5 seconds still to long but better
That seems to be excessive latency giving the number of tiles. Have you looked at Logs (past logs) for errors, warnings and excessive verbose devices?
1 Like
I do have one chatty device the Konnected alarm panel replacement. It doesn't have a UI to shut off debug messages so I'll have to modify the drivers no biggie but that's really it
I have one thought but this may be completely off the deep end, so to speak. I was curious as to what device you are loading the dashboard(s) on as well as is it via the HE app or the Web UI? Is the device(s) you are viewing the dashboard(s) on a wireless device? Could the load latency be a wifi issue be it distance, signal strength, interference, etc? I only as this cause being a network analyst by trade I would questions not only the HE device stats but also what's between it and my viewing device. If you are on a wifi device AND loading via web UI, have you tried maybe a computer/laptop on a wire to see if the loading times improve. this might narrow down where/why there is latency. I know there is a lot of "IF's" in what i said but isn't that most of what we all deal with "IF" this... sorry bad syntax pun!!.. Hope my 2 cent's help!
thought of that but and this is interesting when I use @jpage4500 android app it flys
so the makerapi performs just fine... but dashboards don't
is it only slow on your cell phone? or via a PC web browser as well?
both
but much better since the reboot...
@pat.richards what are you modifying in the Konnected drivers to shut off Debug messages? I am not that knowledgeable yet but am learning from everyone here every chance I get.
Thank you for your help
all I plan to do is comment out any lines like this one
log.debug "$device.label is $stateValue"
I will simply change it to
// log.debug "$device.label is $stateValue"
Make sure your Dashboards don't have any ghost devices (devices that were deleted without first removing them from Dashboards).
Ok learning opportunity 
What is the proper sequence to remove a device from HE? Does it matter?
dashboard, automations, device itself....
First remove the device from any app that it is used in. Then remove the device.
(Dashboard, Rule Machine, MakerAPI etc. etc. are all apps)
6 Likes
If you have not already, try creating a new dashboard and just add one device. Make sure it is the only device "allowed" on that dashboard, and then see how it reacts. If it is still slow I may be a hub issue (too many aps, or an app/devices not running right). If the test dash runs great it could be a tile. At that point I would add tiles a few at a time until you isolate the offending device/app/tile.
Also note you can disable app and device using the gray X in the top right of both apps. After hitting that a check box will appear to let you disable apps/devices (much better than having to remove them all!)
1 Like
This is just my own little sample size here, but I have been watching the board for this too.
I think Maker API is doing something bad. It completely bombed out my hub when I tried this along with the same Android tablet app from jpage4500. I completely removed Maker API and the hub returned to normal.
I have seen quite a few people having slowdowns and one of the common apps they use is Maker API. Funny enough, it doesn't seem to affect those outside apps or devices (like your Android dash experience), but it sure seems to strain the Hubitat itself. Here is just one example: "No response from hub error" on Cloud dashboards
This is just a theory at this point, but might be worth investigating.
1 Like
This would be a nice to have. If we could monitor the resources used by a given app that might be helpfully. Or if there was some way to tell the resources needed by the apps you are running.
Of-course knowledge = danger and all.
I'd buy the Pro version in a heartbeat. Many features like those as well as full POE support have been kicked around the community. For now one can only dream.
I am a very heavy user of maker.api since I use Node-RED. I did have some hub spamming issues - if you hit the server with a lot of requests in a short amount of time things can bog down. Also a flaky zigbee device which sent my zigbee network offline also caused major hub slowness.
I am now running 3 instances of maker all pointing to the same Node-RED server. Have broken things up by Z-Wave/Zigbee + other stuff.
Things seem to be okay. Main hub still very responsive after a couple of weeks. Not sure I actually need separate maker instances but they do help isolate things.
2 Likes
I have 2 instances of MakerAPI on each of my Hubitats (one Node-RED, one Homebridge). Totally stable. Last several months, I've only rebooted when there's a platform update (or an extended power outage).
2 Likes
I know most your logic if not all is done in NR.
Do you have rule machine installed ?
I do - because I occasionally write rules for others (and then delete them once I send the screenshot).
1 Like
While the answer for now appears to be "No", you might want to check out the Hub Watchdog app. You can track the performance of your hub and both radios, observe performance degradation, and even fire a hub reboot automatically. Take a look here.
1 Like
