A Message from Alex on Platform Improvements and Our Plan Forward

I have started having SHM issues again. This morning, my morning rule ran and disarmed my AT&T alarm. As a result, SHM was requested to disarm as well. It didn’t. I noticed and manually requested it to disarm. Took some 10 attempts and app kills and it eventually sticked to the Disarm mode. Took the dogs out and all seemed fine. Until my wife woke up and went out into the patio (same I came in from earlier) - all of a sudden, alarms starting pouring in (door open, patio motion sensor, etc.). Best of all, SHM was still showing Disarmed, even after an app kill. What gives?! To put the cherry on top, my mom’s phone presence suddenly decided to join us. My mom shows present, even though the logs say she last left our home about 5 days ago. Regardless of the logs, she suddenly shows present - this did not happen yesterday and she hasn’t visited us in 5 days. Whatever ST is doing with caching and db separation and log processing, it doesn’t work. Seems like they have a log server that works fine (haven’t lost any logs - all actions are recorded properly) and a SHM server that maintains its own state for the hub - but can’t handle the same amount of traffic the log server can, leading to lags that eventually lead to inconsistencies in current state, hence the I-disarmed-it-but-it’s-still-armed cases.

I bought my hub almost two months ago. I am very close to giving up on ST because… it’s fatally flawed. I can’t rely on my siren to turn on if a sensor is triggered via the cloud. ST needs to get off their horse and acknowledge that their approach to security is not sustainable. A beautiful dream that will always be a dream. They have a growing base of customers/hubs and at the same time, existing customers are adding new devices, so we’re talking about a two-dimensional growth. Whatever they do, they won’t be able to master an ever-increasing amount of events passing through a central system. They need to node-ize the system. Key element here: hubs share nothing in common! That means they can replicate their systems using smaller units that can handle say 100-1000 hubs each. But the best solution is to migrate all logic into the local hub. If the hub can handle 4 video streams, I fail to see how it can’t really just run the groovy code which is perhaps negligible compared to video caching/processing. Decentralize and split the work load onto the client hub. Why would my sensor need to go to the cloud to activate a siren when all three are in the same network? (sensor, hub, siren) This is not a sane security model as their cloud introduces so many extra points of failure. Security Systems must be self-contained, not relying on external components that a possible attacker may hinder (i.e. cut the dsl line on the side of the house before proceeding to breaking in effectively turns SHM off - mean joke follows - much more effective than using the Disarm button - pun intended LOL).

And if a v3 hub is needed to move all the logic locally on the hub, I’ll take that risk and buy it. But at this point, as an embedded/electronics programmer, I feel this battle has been lost. It’s - again - fatally flawed.

SmartThings? Need help? We’re here… many of us are programmers, many are network admins. Use the free and open resources we represent. Please.

Adrian

1 Like

@slagle Just wanted to mention this for a few days now, I have a different problem that many others may share. If I restart my AT&T UVerse router or if AT&T simply goes down and recovers, the ST hub needs a restart as it never seems to reconnect (or at least it does not do so in a few hours). This is the reason I had to remove the batteries from the ST hub permanently, so I can easily restart it. The problem is with UVerse (plethora of bad words and swearing go here), it hijacks the DNS whenever there is no broadband connection, pointing any page you request to their internal IP (by default 192.168.1.254). At this point your browser will think www.google.com is 192.168.1.254 for example. The hub perhaps encounters the same issue, where the ST cloud is suddenly located internally at 192.168.1.254 - ST needs to flush the DNS when trying to reconnect to make sure they can reconnect - this - again - happens with AT&T U-Verse - some smart schmuck at AT&T decided to hijack the DNS to inform the customer the broadband is down. What an #(@#$&@#$^*@#@…

1 Like

If they came out with V3 where everything ran locally, I would definitely buy it. I’m sure a lot of people would and that would most likely take the strain off their database and make v2 more stable.

3 Likes

Unfortunately, this has been a known issue for some time. See the following active thread, which was started 3 months ago:

1 Like

Thank you. I might know the reason it never comes back though :slight_smile: I’ll update that thread.

1 Like

Ha! Thanks JDRoberts for sending me there, after I posted my theory, @mccarbc has found the workaround solution for the restart-after-internet-down for those who have UVerse. It involves disabling an event notification on the UVerse router :slight_smile: Another bug bites the dust - ST could help though by flushing DNS before attempting to reconnect.

5 Likes

Configured it in my U-verse NVG599 Router
Diagnostics > Event Notifications
I un-checked “Broadband Status Notification” and saved. I tested it by unplugging the router. The ST hub light turned Blue as the u-verse Router rebooted, and as soon as service was restored my ST hub was back on-line and the light on the front of the ST hub turned Green… It works !!! … Thanks !!!

3 Likes

My system is working as it should now. I’m just ultimately happy they see what a big issue this caused and are trying to correct the issues.

And since you’re hiring, I specialize in accessibility strategy & implementation :wink:

5 Likes

My SHM still doesn’t work. Worked with support and he (Jared) helped me disable my SHM and set up custom alerts. While helpful I am still not happy that ever since the “fix” Alex talked about, my system does not work as intended. If SHM does not work I would like to remove it from my dashboard and add a working security setup (SmartAlarm) to the dash.

@billyb3

I sympathize with your frustrations. We are still working very hard to get things up to par. We have fixes lined up over the next weeks and months that will help the entire platform.

Keep a lookout for release notes posted by me that will detail all these changes as they roll out. In the coming weeks and months we’ll begin to see meaningful incremental improvements.

1 Like

Send me an email with your CV

Slagle@smartthings.com

1 Like

Then why the big pat on the back and false sense of relief for this user base?

1 Like

Yeah I have to say I am not thrilled with them setting a status of resolved on the SHM issue. Especially when their “fix” went in when they broke mine.

1 Like

I have to say my SHM has behaved since the update. It was broken before for sure.

SHM was broken for me before the update, now it’s broken differently after. The state of SHM is not represented accurately in the app, and random things are not working (lights not coming on at sunset, routines not triggering automatically, etc.) I’ve done support tickets, but they mostly just say it’s a known issue and move on to the next,

2 Likes

We bought the SmartThings system and devices last November as an alarm system because our house was broken into one night and the criminals made off with all of our valuables on the first floor while we slept on the second floor. I bought this for the safety of my family.

I could have gotten a dedicated system from an alarm company, but I thought that with a home automation system it was an alarm system plus more, and that we could build upon it.

It has never worked correctly. I don’t mean we’ve had stability problems recently, I mean it’s never worked correctly.The door sensors don’t report open/close correctly… the alarm is triggered, but the siren doesn’t go off… or the siren goes off when the system is in disarm mode and I open a door. I have never once been able to turn off the siren with the app. Even the initial setup didn’t work right. It takes forever, and several tries, for the system to recognize new hardware.

This makes me angry because it makes me feel stupid. That I didn’t do enough research. That I was too trusting that the problems were temporary and would be addressed. That I let this go too long without demanding a refund. I feel stupid because I told my wife… “no, we should get this SmartThings system instead.” Now I’m out a few more hundred dollars with nothing to show for it but frustration.

I am shocked at the design of the system. I can’t understand why it was decided to build a home automation system that sends all of the logic and instructions through some remote mother system. This is home automation… it should be built to fail-safe, not fail-first. If it can’t connect to an API it should still be able to operate. It shouldn’t go haywire because you can’t scale your noSQL system… or because of system load.

I really just want this system to work. I would still want the functionality if I could rely on it. But I have no confidence you’ll be able to turn this around. It’s not just a couple of glitches. So barring that, I just want my money back.

3 Likes

I was able to get my SHM functioning properly again by deleting ALL of it (security, leaks, smoke, etc.) and deleting any arming/disarming from routines. I then set it all back up and it has been stable for 3-4 days. Now whether this actually is repeatable by others, or I just got lucky, I don’t know.

Bob, I totally feel for you. You shouldn’t feel stupid. The SmartThings marketing machine is strong, and it presents a glossy view of what is a rocky user experience. Reviews in online and print publications are similarly superficial. Unless you had come to this forum and read through the thousands of accumulated posts, you never would have known.

You should email support and ask for a refund. It is never too late for SmartThings to do the right thing. If they won’t issue you one, send a private message or email to Alex and Ben.

1 Like

Just tried this while at work. My app finally updated with the correct SHM status. The test is when I get home for the “I’m Back” routine. Thanks.

1 Like