Anyone else having issues with the Smart Home Monitor after the update? No matter in which mode I am or how many times I disarm it, I still get intruder alerts.
Thatās prime development time for me, and Iām on the east coast!
Seriously, though, unless/until the platform gets to a state where they donāt need downtime for updates, this is a damned if you do, damned if you donāt kind of thing. 4am EST/1am PST is probably as good as youāre going to get.
The length of this one just made it too long to fit in anyoneās window of complete comfort.
1 Like
tgauchat
(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy)
24
This .
When was the last time you couldnāt login to Amazon at any minute of any day and buy something with one click (including SmartThings, ironically, if it was in stockā¦). We live in a 24x7x365.25 world. So should our Smart Homes.
Ultimately no matter what you do, no matter how much of a 24/7 company you are, no matter how many things you put in placeā¦ at some point youāre going to need downtime of sorts.
There is always downtime, sometimes you just donāt notice it. Takes a lot of engineering. Iām not sure ST is ready for that, but it sure would be nice.
That being said, more local control = less consequences, Iād take that too. In the end, still happy with my experience.
1 Like
tgauchat
(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy)
27
Technically, yesā¦ But waitā¦Google has achieved 99.798% uptime with ZERO scheduled downtime for maintenance,ā¦ So, no ā Downtime for maintenance is not inevitable.
And yetā¦ theyāve had downtimeā¦planned or not, minuscule or not, theyāve had downtime and thatās a company that has a multi billion dollar budgetā¦ and they still get downtime.
It also entirely depends on the nature of the business and what services they provide, just because both Smarthings and Google are expected to be 24/7 companies, thatās where the similarities end, Smartthings and Google provide different 24/7 services.
Hereās the thing, weāre a 24/7 company, we strive for zero downtime and but itās impossible to achieve ZERO downtime and in our sales contracts we actually have a written agreement that says we are allowed 4 hours of downtime for scheduled maintenances every month. For the most part we donāt need it however, there are isolated scenarios where we do and we need the clause to prevent us from being sued. I will place money that Google has a similar clause in their contracts
My bank frequently has a banner when logging in of upcoming maintenance and downtime.
Ideal? No. But I canāt say our services that we host at work are up 100% of the time either, and Iām sure many of us have lived through some harrowing upgrades.
I never received the follow-up notices that were mentioned here, only the email last week. It doesnāt matter though, as I had zero issues during the maintenance window or since.
I thought I had, but upon checking just now, it is offering me the option to subscribe. Guess I got distracted at the time. Thanks for the reminder!
1 Like
tgauchat
(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy)
34
I agree that context is everything. I think SmartThings needs to aim hit the level of a Utility, though local processing will buy them leeway.
In other words, most people in the USA would not tolerate 4 hours (or even 1 hour) a month outage of their electricity, cell phone, or Internet. I donāt know what the statistical outage rates are for those, especially distinguishing planned from unplanned.
Meanwhile if Googleās infrastructure helps them (Nest) create a 99.8% uptime service, this will be attractive to people keeping track or anyone who just happens to be impacted too often by inconveniently scheduled or unscheduled outages of a competitor.
Oh absolutely, they SHOULD aim for zero down time! Heck I hope ST sets such a precedent for all other HA companies that the government starts to consider it a utility but that would be a ways off.
Ironically, I donāt know about you but where Iām at, we frequently have power outages at the beginning of the year, especially when there is thunder and lightning, and yes, for hours at a time on occasion. Hell, they even love to over-volt on occasion and blow up my devicesā¦
At the moment I still see Nest and Homekit as cute little attempts at home automation but are highly restrictive in what they can do, heck Iād even go back to Wink before I considered those two options and we all know how I feel about the hatefulness that is the Wink Hubā¦
Personally, Iād just like it to be as reliable as a dishwasher.
For $300 (the price of a starter kit), it doesnāt have to be perfect. It will fail or work imperfectly every once in awhile. But those times should be rare enough that I donāt remember exactly when the last one happened, and certainly no more often than 2 or 3 times a year.
Different People May Have Different Expectations
On the other hand, I personally donāt use SmartThings as a security system. (I have a completely separate setup for that.) My feeling is that if something monitors potentially life and death emergencies, it needs to be as reliable as a smoke detector no matter what it costs, which means 5 years of near perfect uninterrupted operation with a very small amount of allowable false positives and no missed alerts.
(Thatās why Iām ok with security systems that charge a monthly fee. And why Iām ok with them limiting coverage to a few specific device models and limiting custom code to a few IFTTT type rules. I know Iām asking a lot more of them than I do of my dishwasher, and I know itās challenging to meet those reliability targets.)
If I had bought SmartThings to use as a security monitor, I would have a much higher level of expectation for reliability.
But regardless of which category SmartThings belongs in, it should have a predictable level of reliability.
Right now a bunch of stuff that worked last week isnāt working for me. Thatās on two sides of the platform update, but I have no idea if thatās related. Just weirdness cropping up. Some of it comes and goes. For me, thatās the hardest part of knowing how to plan. Announcements of planned outages are helpful, but I never have any idea of how big the " ripple effect" will be after it comes back.
The following is from the afternoon after the maintenance was complete.
The switch was neither On nor Off. Results were, as they say, āunpredictable.ā
( and, no, this was not just a mobile app refresh issue. The reason I was looking at the screen in the mobile app to begin with was because the switch was not working properly. )
This is a bit odd. My time-based auto-routines still did not run this morning. Itās almost like the maintenance killed the scheduler for those as well. I guess Iāll try and just click through them to see if that jumpstarts them. Regardless of the maintenance window, this is not the kind of thing that should persist after an update.
Edit: I looked at the details in graph.api, and things are even stranger. It claims the routine ran at 5:00am like it was supposed to, but it most certainly did not. The mode never changed, and no entry was created in the messages screen.