So, whatās the general consensus from all of you that had the severe load issue? Has .138 (and subsequently .139) resolved the issues? Iām still sitting on .128 and havenāt upgraded because of these issues.
I've been on .138 since it dropped and I have not any load issues like I was before
1 Like
I haven't bothered to keep the hub running long enough to see if I got the message with .138 or .139, mainly because if the lights don't turn on in the morning when the wife gets up, then that opens up uncomfortable conversations as to why I bothered buying all this. For the version or two before that I got the message a couple of times before I just start pro-actively rebooting every day or two. With .138 I noticed memory being much lower than I recall previously, and after about a day, I'd start to see occasional spikes in CPU which would increase in frequency gradually. Not enough yet to trigger a message, but enough to convince me to reboot. We'll see if .139 is better. I will say that .139 started with more memory following its reboot than at any time with .138. So fingers crossed it's much better for me. CPU looks good, but it hasn't even been 3 hours yet so we'll see.
2 Likes
It has been solid for me on .138 for 4-5 days so far. Not sure if it was the software or my removal of HomeKit Integration or both but no issues. I will (eventually) test adding HomeKit back in to see if that changes anything.
1 Like
I can confirm also that the latest platform updates solved the elevated and severe hub load issues.
1 Like
The only issue Iāve found with .139 is it breaks the Zen thermostat.
Iām still parked on 2.3.3.140 and wonāt be moving until thereās a definitive āfixedā on this, ideally with a little explanation as to what was so broken.
Three completely borked hubs and the lights not working does indeed raise uncomfortable questions as to the suitability of my purchase, and rightly so.
2 Likes
Huh, okay so the word is that this is "fixed".
Fixed high CPU usage/memory leak under certain conditions.
@gopher.ny Is there any chance we could get some detail on what these certain conditions were? It would rather restore my confidence. At the moment this platform-disrupting bug is just subject to mystery and intruige, two things I don't like powering my home. 
I had the issue once on 2.3.4.137, then with 2.3.4.138 logging was added
Ive had no issues since .139 either
Lets see what gopher.ny has to say on this
1 Like
An explanation wouldn't make sense without knowledge of the internals of hub software.
6 Likes
Even if we donāt know the internal details, it would still be nice to get a general explanation as to the cause and maybe what is planned to avoid it in the future. At the least it would go a ways in helping with trust in the platform.
6 Likes
In general, I'm sure that something was doing something that it wasn't supposed to do. They stopped the thing from doing the thing it wasn't supposed to do. Now they know to look out for the thing doing that thing it wasn't supposed to be doing.
Ha! I should take up doing stand-up.
9 Likes
Spot on response. As a new HE user, the reply left by @gopher.ny comes off as dismissive and doesn't give me a lot of confidence in HE's commitment to clear communication or transparency. I made the time-consuming move from ST in the fall because I felt Samsung had completely failed their user community in communicating key dates and specifics around the Groovy EOL. As they say, "Things that make you go hmmm."
4 Likes
Honestly, this is the first time in my several years of ownership several that Hubitat had a major performance issue to deal with.
4 Likes
I can understand pause for thought having just moved onto the platform and experiencing performance issues such as this. I would encourage you to spend some more time taking in the experience over a slightly longer period, seeing the benefits it can bring.
I'd also add that while it's not ideal when issues like this come up, I still feel that the response time is commendable for the HE engineers to identify the root cause and get a fix out, also given the time of year. I can only imagine there are situations where diagnosing issues like this can be tough and require multiple iterations of development and testing across a number varying setups.
Personally I am happy to leave the details to those who continue to deliver us a great experience. If there are a few glitches every now and again, I can accommodate that, and I'd prefer to judge them on their record over a period of time, rather than pinning them down on the details of a single issue.
9 Likes
That and the ability to roll back to a previous stable version for those who need it, is what makes it ok for me.
Although Iām interested in details (from a personal perspective) I understand that this is not always relevant for all of the community. The options we have on providing feedback and working around a given issue is what counts for me.
4 Likes
Really? I started this thread back on 27th November and it's taken two months to seemingly get to the bottom of a platform-breaking bug. I've had very little interaction with the HE guys (nothing beyond what you see on this thread) and it would just be nice to have a little more than "trust me, bro - you wouldn't understand" as the feedback.
That's the ticket.
I know not everyone here is a sysadmin or a developer or a Linux nerd, I just happen to be all of those things and a note that said "ah, it was issues with the way we did Thing X for Feature Y so we've applied Fix Z to it" would be appreciated. At least that way if I, or anyone else, sees similar issues again we might find this thread and know to try turning Feature Y off. Or even know that there's absolutely nothing we can do and should report it quick-smart before others can't control their lights.
I'm not looking for any beef here and really very much appreciate the work of the devs who have found the issue and smushed it. Once I've seen enough positive feedback that I'm happy to update and test (as I'll need to do all three hubs at the same time) then I shall nip back and mark this as solved. Cheers, all!
5 Likes
All you need to do is either A/ raise a support ticket, or B/ reach out to staff like @gopher.ny or @bobbyD . Iāve always found them to be helpful.
I got access to beta versions within days of providing my hub ID to Hubitat and giving them permission to scrape engineering logs.
You may have had a unique issue, as most of us here didnāt run into issues until after the HomeKit underpinnings were rolled out.
Yeah, I did. Provided my IDs and all that jazz. No dice. 
This was from the first 2.3.4 release with the HomeKit integration. I'm a Homebridge user, which is why I dived in to see how this fared in comparison. The HomeKit integration worked great, but thank goodness for the rollback feature, saved my bacon! 
