Missed motion events; outage?

#61

Awesome

To confirm what you are saying, my test piston that plays ping-pong with IFTTT showed 66 outs and 67 ins this morning. I’m not sure why I ended up with one extra in than out, but at least it isn’t 20% loss like ping-pongs through smartthings.

As am I. Are we thinking that there are multiple servers, and that might explain why only a handful of us are having the issues? Another friend of mine is not having these issues. Is that a luck-of-the-draw thing do you think, or can I simply move to a different shard in the cluster?

#62

i’m on na04-useast2 as well and also experiencing the problems described above.

#63

I’m on shard 1 and have seen it too.

#64

Getting hit hard right now:

org.springframework.jdbc.UncategorizedSQLException: Hibernate operation: could not execute query; uncategorized SQLException for SQL [select this_.id as id52_0_, this_.version as version52_0_, this_.date_created as date3_52_0_, this_.key as key4_52_0_, this_.type as type52_0_, this_.value as value52_0_ from server_config this_ where this_.key=?]; SQL state [null]; error code [0]; [SimpleAsyncTaskExecutor-467] Timeout: Pool empty. Unable to fetch a connection in 30 seconds, none available[size:50; busy:1; idle:0; lastwait:30000].; nested exception is org.apache.tomcat.jdbc.pool.PoolExhaustedException: [SimpleAsyncTaskExecutor-467] Timeout: Pool empty. Unable to fetch a connection in 30 seconds, none available[size:50; busy:1; idle:0; lastwait:30000]. @line 1149 (processSchedules)

#65

I just wanted to add my name to the list. I have also recently started getting missed events on my motion sensors. I don’t know exactly when it started as I was away from home for several weeks but it started happening as soon as i returned (which was Dec 26th). Sometimes the turn light on misses, sometimes the turn light off misses.

I am using the Iris motion sensors with the “SmartSense Motion Sensor” device handler. They have been working fine for a year and just started acting up upon my return. Batteries all register 67% or better on my 3 sensors.

I see multiple threads on here probably discussing a similar issue…

I modified the trigger slightly to be a “motion changes” instead of “motion is active” but that does not fix it. Here is a sample of one of my motion triggers:

2020-01-02%2010_46_46-webCoRE%20Dashboard
2020-01-02 10_46_46-webCoRE Dashboard.jpg770×818 162 KB

#66

I created a simple trigger within smarthings to turn on my shower light (use it as a night light) and disabled my webcore version so I could test within smarthings. Smartthings itself was failing to turn my light on even though the trigger was sent to the light.

So in my case, it may not have anything to do with the interactions between SM and Webcore

Here are the triggers in the log, but the actual light did not turn on 2 out of 3 tries.

2020-01-02%2011_38_51-Device%20States%20List
2020-01-02 11_38_51-Device States List.jpg1924×275 54.7 KB

#67

Yeah, most of my issues are pistons not recognizing that virtual switches have changed. I do everything around switches so that I can easily toggle them in ActionTiles. So I’ll see my “Downstairs Presence” switch flip on, but the piston that either has it configured as a “changes to on” or an “is on” with “Always subscribe” does not trigger reliably (maybe 60% of the time).

#68

I am also on na04-useast2 and having issues with motion sensors and cell phone presence detections. Everything shows up fine in smartthings “recent” activity, but many (not all) dont show up in the webcore pistons log. :frowning:

#69

Chiming in with the same issue here. I noticed the outage on the 19th and had some issues the next day, but then was gone for a week and have had numerous issues since coming back. Most notably motion sensor tied to a light not shutting off, like a lot of others. Everything logs as expected in ST, but not in WebCoRE. Also on na04-useast2.

#70

Just want to share, that I have been experiencing the same with the most simple pistons, as the one below. Sometimes it only catches the “on” trigger, sometimes, only the “off” trigger. Also on na04-useast2. All events are being captured by ST, but not WC… :frowning:

fireplace-piston
fireplace-piston.png1178×668 77.2 KB

#71

Do you know they are being “caught” by SmartThings (ie a trigger set up in an ST automation not firing)? I’m setting up a test on a trigger I know my pistons have been missing, but I’ve read others have set up a similar test and though ST recognized the event (motion/switch state) it doesn’t trigger the relevant automation in ST.

#72

All my triggers are being caught by ST and notifications are being fired when I setup similar automations in the ST app.

#73

Ugh, yep. Just confirmed on a trigger that ST detected that WC did not. :sob:

#74

@WCmore, anything we can provide that would be helpful? Doesn’t look like logs are going to help in this regard since the trigger is not detected by WC.

#75

I have nothing new to add, because I think my advice above is still relevant.

Remember, one sloppy (or careless) piston, can easily impact a simple (well written) piston… and of course, if multiple pistons are trying to execute when a device changes to X, then sometimes, one of those pistons will not receive the trigger.

Your mileage may vary.

#76

@WCmore, apologies, I thought you were a developer with WebCore. Thank you for your suggestions but I do not believe they apply here. Multiple people reporting the same issues even with virtual switches starting at the same time does not feel like the same old “try to be more efficient” situation.

#77

I am not a dev for webCoRE, but a hard-core user (and programmer) with tens of thousands of pistons under my belt.

Just to clarify, I am not defending SmartThings here, but I was trying to emphasize how our code impacts our entire SmartHome.

IE: When too many things are trying to happen in any one moment, occasionally one (or more) may get lost in all the commotion. I have seen this hundreds of times!

#78

So, @WCmore what would be the recommended best practice if you need to perform different actions based on 3 different states of the same device? Say, I need to get a notification when my garage door changed to open, then a notification when the garage is closed and also a notification in case the garage is open for XX minutes. Would creating 3 different pistons, one for each trigger, help? Currently I have one piston with 2 triggers (changes to open / changes to closed) and another different piston checking “stays open for XX minutes”… Thank you!

#79

All three of these use the same trigger. (specifically: GarageDoor’s contact)
Whenever that contact changes, any piston with those IFs will execute.

Because of this, I would try to put all three in the same piston.
(and then consider it a single trigger piston)

Pro Tip:

“stays open for XX minutes” will trigger when the door opens or closes, but will not do any action until the timer expires.

Because of this, that block can remain separate, if you like.
(just be aware that with this method, 2 pistons fire at each contact change)

#80

Ugh, that’s frustrating. I just split many of my “Switch” pistons that had both “On” and “Off” activities on them because I thought maybe it was the complex pistons that were causing my conditions not to trigger. I had also added two motion sensors to relatively high traffic areas so I was rewriting my logic to pause pistons that utilized triggers on these motion sensors when motion was detected in an area. The piston would then be resumed when motion was detected in another area (and those pistons would be paused and so on). Unfortunately this appears to already not be working as smooth as needed for the same reasons as why I’m exploring it in the first place, because pistons aren’t triggering or having to be run twice.