I note again suspicions about the Hue app. (but I'm just deducing from my observations on here)
Does anyone who experiences slowdowns have the hue app on that hub?
Do you get slowdowns on hubs that dont have the Hue app installed?
I only have one hub (C-5), and I have the Hue app installed. I only have Hue smart bulbs, and they're all via the Hue bridge and not visible to HE.
I have two Hue motion detectors: one attached to HE, and the other only on the Hue bridge.
ETA: I am getting slow downs. I've made some changes in the last few days, and now I'm waiting to see if the slowdowns continue. One of my RM4 rules was misbehaving, but I think I've resolved that issue. It is a rule that automates my kitchen lights, which are Hue. Whether the rule was the issue, or Hue is, remains to be seen.
One C4 here with the hue app / hue hub. No slowdowns.
I do, and have always suspected Hue., but no proof. Although on my other hub, I have very few apps, only hardware. I think I’ll set up some simple rules to Ikea bulbs only which are attached to that hub and compare the two.
2 Likes
Yup and I have thought of getting rid of my hue lights and hub.
I have the Hue app with 3 RGBW light strips
I was getting slow downs but this turned out to be caused by Chromecast (beta). As soon as I disabled it things went back to normal.
The disappointing part is this was back in May and there has been no update to the app that I have seen in the release notes for the updates since then.
(I have got by without it so its no big deal for me).
So I read this entire thread and have very mixed feelings.
As many of you know, I am not a developer, have stated this many times, however I am what I would call pretty tech savvy.
I find that Simple Lighting at times to be too simple, when I asked (long time ago) to have a time feature added (IIRC), I was told that was outside the scope of SL, it's simple for a reason. So what am I left with but RM, and again many of you know, some of the logic/terminology doesn't doesn't translate in my brain.
Notifications, same thing, very simple yet the only other option is RM, back to the issue above
Seems to me, this system is for the beginner (SL, Notification...) which appears to be rock solid and very advanced via RM which IMO is like walking through quicksand or customer code which is way more dangerous. I found out the hard way, wanted some way to make audio announcements on my Alexa, used Echo Speaks and it wreaked havoc until I removed it. The new beta doesn't seem to be affecting my system, however for me, it's just not very stable at the moment
I have removed everything except for Alexa Speaks - Beta and Home Tracker, that's it and at this time, my system runs very smooth and fast
Thanks for listening 
1 Like
Maybe RM could pop an error message in the log if it detects a rule thats thrashing and taking up too many resources. ie if it loops or fires more than $threshold times per second, or if its taking up too much memory.
Same with the sandbox that user apps and drivers run in.
1 Like
Both of these would be great! I would buy another hub if that were the case. First because there is a built in way to sync devices to all hubs then. And second I would absolutely put all 3rd party apps on their own hub...and then hopefully have a way to see which one is behaving badly.
The ONLY gotcha to this plan is custom drivers. if you are going to have a dedicated Zwave or Zigbee hub...you'll probably end up with at least one or two custom drivers in there. However even with that...I love the above two ideas.
I've been lurking since this post was started. It's hard to believe I'll make post #190 in less than a week. Obviously this is a topic of concern to at least a vocal subset of the Hubitat community. I have a background in software products but have been more on the IBM and Microsoft stacks so I don't know the internals of the hub and what's possible. That won't stop me from making a couple of comments, 
. It seems like this thread is degenerating and starting to go back and forth with philosophical and devil's advocate arguments. There are still some concrete ideas though. I hope these two thoughts fall in the latter group:
- I like the idea of the community providing feedback on apps but it has to be both negative and positive. It would be great if each community member could list the applications they have used and score it from -3 to +3. The votes could be consolidated and viewable by the community and the developer of the app/driver. It seems to me we'd get a better sense of things if we had people list all the apps they use in one place rather than depend on comments to the threads of each individual app. People tend to post negative comments rather than positive ones so it's hard to get a sense from the threads how many people are using an app or the balance of positive/negative. Having one place to "vote" could help with that. Companies use the Net Promoter Score to get feedback and this would be similar. I know Echo Speaks has been identified by some as a problem but I haven't seen an impact from it. I know @tonesto7 has spent a huge amount of time with V3. It would useful to know if problems that were previously identified have been resolved.
- Is it impossible to get any instrumentation to help with these issues? I trust @bravenel's comment that CPU is not typically the bottleneck. Is there a way to determine "I/O" and network request counts or elapsed times for certain requests that can be problematic. Could rule machine count "lines" executed or whatever it is that can impact the hub's performance? If some statistics could be captured and viewed or maybe checkpointed as part of the nightly backup, people might be able to see a) anomalies in particular apps/devices/rules b) changes from before and after custom code was added/changed.
I like the fact that Hubitat is open and extensible and that there is a vibrant community to add applications and drivers. I like that privacy is part of the design and what happens on my hub generally stays on my hub. It's impractical for the Hubitat team to support or police what people are doing and I wouldn't want that anyway. However, empowering the community with some easy way to identify problems so they can be fixed (rather than just complaining) and having some high-level metrics that I can use to monitor issues would be helpful. Just my $0.02.
6 Likes
Ranking system, I would say no... how long until people are not voting on stability, but on it doesn’t do what I want, how I want. I couldn’t get it to do what I wanted... it wasn’t easy for me to use... etc. looking at amazon/other online ranking systems, not long. It is too easy for mis-information to get into this.
4 Likes
On the responsibility of HE, why is SW not given the same as HW. If a user put a 19v dc adapter on their hub, would we say HE is responsible to keep someone from doing that. Yes the box says it supports user development, do they need to put on the box “responsible user development”?
1 Like
True, but I am losing zero sleep on malformed user zwave and zigbee drivers. Yes, they COULD cause issues, but in my experience it hasn't been a big problem (for me). I write all of my own zwave drivers (for anything that I can't use the in-box driver for), so I know what they do. And I at least look at the zigbee user drivers I use to make sure I don't see anything really dumb in there. I have my zigbee sniffer working finally, so may make a swipe at making some zigbee drivers in the future, too.
I can live with the user drivers - for zigbee and zwave at least. When it comes to LAN drivers, well that is much dicier, but those can go on another hub if segregation were made really easy.
This was what Vera tried to do as well... didn't work out so well as the master hub still needed the app installed to actually know what the data was. Their implementation was basic and worked well for hardware devices to extend the z-wave reach and network node limits but it was terrible for apps. Then the OpenLuup project started in order to do a full hub offload of apps. This worked well and was a community project.... this all seems very familiar.....
1 Like
Removing stuff should be irrelevant for me. I've been fine for months and the last few weeks this has started. It's to the point where I'm ready to unplug it completely
This will depend on the apps or drivers used. It could be cloud issue with the product and the app is not handling it correctly.
Sorry for the troubles. I don't recommend "removing stuff", we understand it takes a lot of effort to set up a system just the way you want it. But when things go sideways, I do recommend disabling at least the last custom code you have added to the hub, to see if that is causing the problem. If you didn't add any new apps or devices, and you are not running some of the apps and drivers that are known to cause slowdowns in some hubs (but NOT all), such as Chromecast, Lifx, Echo Speaks, etc then please reach out to support@hubitat.com. We would be more than happy to take a deeper look at your hub to see what else might be causing the problems you are currently experiencing.
If you don't know how to disable an app or a driver, please see "Disable Device Drivers" and "Disable Apps" sub-headings in the following documents:
https://docs.hubitat.com/index.php?title=Devices
https://docs.hubitat.com/index.php?title=Apps
Well, I finally broke down and deleted Echo Speaks. I was really hoping this new version would solve the hub slowdowns the old one was blamed for, but it does not. I could actually look in the logs and watch the cookie validation times increase throughout the day from 330ms to 2100 ms ( then reboot) back to 330ms and creep back up to 2000ms over a day and a half, so I knew it had to go. I put it on SmartThings, where it works great, and am using HubConnect to enable automations on HE. I have disabled echo speaks in the past in my search for what was slowing down my hub, but still experienced hub lockups and slowdowns with zigbee devices and the web ui. I didn't realize that NST manager was perhaps as much to blame for the problems. I deleted NST manager a couple weeks ago and noticed an improvement, but after deleting Echo Speaks the change is night and day. The web ui is amazingly fast (I had no idea that it could just snap between pages without any real load time), and zigbee devices respond the instant I push a button. BTW, there was also the netgear issue, which helped me as well, because my hub didn't respond this fast when it was brand new.
Thank you to the developers (and shepherds) of HubConnect and Homebridge-Hubitat-HubConnect: @srwhite, @csteele, and @dan.t for making great apps that allow the hub to really shine! Also thank you to @bobbyD, @bravenel, and @mike.maxwell for your commitment to being involved and responsive to the needs of the community and making a great product. Now that this problem is fixed, maybe I can figure out how to get mode changes to sync with Homekit.
6 Likes
Both Echo Speaks and NST manager are from the same author.
We're glad to hear that removing these two apps has resolved your hub slowdown problem.
1 Like
No offense to @tonesto7 and all of the work he has done on this, as his apps always worked well for me on SmartThings, and now continue to with HubConnect. I should have uploaded a screenshot of the logs to him, but the issue wasn't errors that were happening, it was just logging that showed the 20-minute check-ins to validate the cookie that tipped me off. I brought up a page of these logs by themselves and saw that as the validation times got longer, the hub got slower and slower until it was almost to slow to navigate to the reboot page.
1 Like
