I have the same issue, historical switch items are always OFF. Also see my post here:
https://community.openhab.org/t/oh3-rule-check-in-a-rule-if-item-was-on-in-past-30-minutes/
I have the same issue, historical switch items are always OFF. Also see my post here:
https://community.openhab.org/t/oh3-rule-check-in-a-rule-if-item-was-on-in-past-30-minutes/
SOLVED: I went back through my notes and realized that I had the same problem when I updated from 2.4 to 2.5 wherein I was unable to use home.myopenhab.org to access my dashboards. The root cause of the problem is that I am running a high core count (32 core) server and Jetty chokes. The solution is to either pin the Docker to use only 4 cores or modify the services/runtime.cfg. I had made the modifications to my 2.5.9 installation but OH3 changes from “eclipse.smarthome” to “openhab” and those changes have to be done manually.
For future reference: in Openhab 3.0 org.eclipse.smarthome.webclient has changed to org.openhab.webclient. So it becomes:
org.eclipse.smarthome.webclient:minThreadsShared=20
org.eclipse.smarthome.webclient:maxThreadsShared=40
org.eclipse.smarthome.webclient:minThreadsCustom=15
org.eclipse.smarthome.webclient:maxThreadsCustom=40
Thanks to @mhilbush and @arjanmels for the fix as now my OH3 is running without any issues.
UPDATE: My issue is definitely a Docker problem unique to my installation. I haven’t yet resolved it, but I was able to install OH3 on my Server using a VM and it worked fine, so not a network or OH3 problem, but a local docker install issue. Thanks again for responses.
My apologies. My shorthand seemed to create additional and unnecessary confusion. I am using Home = home.myopenhab.org as that is the Tab on the website. I also have no doubt others are able to access OH3 from home.myopenhab.org, but every time I try I get
openHAB connection error: HttpClient@4016d1a5{FAILED} is stopped
I just created a fresh Docker and did a clean install and I get the same error. myopenhab.org shows that I am ONLINE and I am able to send messages to my iOS device but I am not able to access the dashboard. I will continue to evaluate, but I am curious as to whether you are using a Docker install or another type of setup such as a Pi? I know it shouldn’t make a difference but I am trying to eliminate possible causes.
Which Docker image? Alpine or Debian? The default is Debian. I tried the Alpine image with H2 and moved to Debian due to issues.
I use debian.
On OsX 10.15.17 Safari Pages -> Add from Model wont add widgets, instead just returns back to the pages overview screen. Firefox 84 on the same system works flawlessly.
What about GPIO binding in OH3
This is at least equivalent:
Interesting: the aforementioned latency only exists if the Android client (beta or release) is logged in to the OH instance. If I do NOT log in the client to the OH instance, the latency performance is effectively zero.
Suspecting some kind of DNS-induced wait-states (due to the use of a locally DNS-named OH instance e.g. openhab.mylan.tld), I checked, and discovered that I see the same performance issues whether the client is configured via a DNS name or its IPv4 address.
This issue persists even after the update to 3.0.0.Release.
I’m happy to help reproduce/debug with some guidance.
In my environment the log4j2.xml file was ok, but in the file org.ops4j.pax.logging.cfg I had to change everything from stmarthome to openhab. Now I get the events in the events.log file.
Actually that file should only contain one line:
Configuration is done through log4j2.xml
.
I installed OH3 over my working OH2.5 and the file org.ops4j.pax.logging.cfg has 145 lines. I will try to change everything to only 1 line
org.ops4j.pax.logging.log4j2.config.file=${karaf.etc}/log4j2.xml
Make sure that file has the correct content:
It is exactly that content. I change now the org.ops4j.pax.logging.cfg to only one line and it is working.
Good, now you are on the openHAB3 track
I’m on OH3 (release) and these messages appear not only 33 times as in your log, but in mine I have 64 entries of this. So yes this happens in a clean install too! It’s a bit anoying seing that - any news on this?