Hi all, we are getting ready to move infrastructure providers this weekend for our myopenHAB cloud service. This should not result in any change to our service offering or functionality for you, and there are no changes required on your end.
The cutover should be very quick, i will stop the current service, change DNS to point to the new cluster, export/copy/restore our database on to the new database server, and start things back up on the new service. I am hoping this take 10 mins or less, i’m not expecting any major issues as i have been testing this all week.
Our current provider Linode was recently purchased by Akamai and have started to increase our hosting costs. We are moving to a new EU based provider that will provide us even faster hardware but at a lower cost then what we are currently spending.
I am planning on doing this move Friday, Feb 19th 20th at 11PM UTC, and if for some reason that does not work, i will plan again for Saturday at 11PM UTC. I will post an update when i start, finish or if i need to postpone.
Ok, all done. The actual cut over only took about 10 mins, but it took me a lot longer to track down our ulimit issues with nginx (and tweak worker limits) , this was also borking our health checks that we use for DNS liveliness. Right now i actually have the service slightly scaled down from it’s full potential to measure load and make sure we don’t over do it on containers per host, but so far, even with the “thundering heard” reconnecting , performance seem to be better then before, so thats encouraging !
It was while he was doing the job. The test cert was pushed on the prod website. It turn out in open openhab at that moment and had a warning that the cert didn’t match the domain. Cert was signed only for .us domain. It was fixed about 5 to 10 min after
I’m noticing that my openHAB service running on Ubuntu keeps disconnecting from the cloud at myopenhab.org over the last few days. I think the timing aligns with this migration. Restarting the service allows it to reconnect to the cloud for a few hours until the issue happens again.
I don’t know if this is because I am running on an old version of openHAB (3.0) or if it is because of this migration, but I’m open to any debugging steps/providing specific logs to help understand why this is happening.
openhab.log.4.gz:2026-02-20 16:00:42.913 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.4.gz:2026-02-20 16:26:56.652 [INFO ] [io.openhabcloud.internal.CloudClient] - Connected to the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.4.gz:2026-02-20 16:28:22.689 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.4.gz:2026-02-20 16:29:18.019 [INFO ] [io.openhabcloud.internal.CloudClient] - Connected to the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.4.gz:2026-02-20 16:41:28.592 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.5.gz:2026-02-20 17:41:40.465 [INFO ] [io.openhabcloud.internal.CloudClient] - Connected to the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.5.gz:2026-02-21 00:57:48.654 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.6.gz:2026-02-21 00:59:15.518 [INFO ] [io.openhabcloud.internal.CloudClient] - Connected to the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.6.gz:2026-02-22 10:10:11.147 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.7.gz:2026-02-22 10:39:10.100 [INFO ] [io.openhabcloud.internal.CloudClient] - Connected to the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
openhab.log.7.gz:2026-02-22 16:39:40.020 [INFO ] [io.openhabcloud.internal.CloudClient] - Disconnected from the openHAB Cloud service (UUID = 06...3f, base URL = http://localhost:8080)
We don’t test with versions that old, but i’m not aware of any changes that would prevent older versions from working. Did you try restarting openHAB? If you want to send me a private message with your openHAB UUID i can take a look at the logs and see if there’s anything there.
Yes, I’m basically having to restart openHAB every time it goes offline on myopenhab.org. Each restart brings it back online for a few hours.
Looks like I cannot send a private message on the community yet, likely because I just registered today. Let me know if there is another way for me to share the UUID. Thanks!
Yeah, i just re-read your post, and saw you mentioned restarting . I had made a few deployments yesterday and today, which would have disconnected openHAB’s. Those should reconnect in the background immediately, but i’m guessing your 3.0 version is not handling that well. There were many changes made to the reconnect logic over the years since 3.0. The likely only long term fix will be to upgrade to our stable release.
I updated your forum account so you should be able to pm me the UUID so i can try and look at that.
Try restarting so i can look at incoming logs. Also looking your logs again , there are some odd sequences, like a disconnect at “16:41” and then re reconnect at “17:41” , then later a connect at “10:39” and disconnect at “16:39”, it’s odd to have round hours like that…
So it looks like you have been connected for about 10 hours, i’m going to assume there’s something our 3.0 client does not like when we gracefully shut down a server (it sends a disconnect message with a reason vs just terminating the socket like it used too). The good news is we don’t deploy that often, but at least for the next few weeks, i will probably will deploy a couple times as i just rewrote the whole stack (with claude) in typescript and will likely push a few improvements and fixes. Long term, i would upgrade to our latest openHAB version as that has proven reconnect logic in it.
It appeared on the pop-up warning. I rejected the cert, but the message hasn’t reappeared.
I tried restarting the app and turning off phone wifi to force connection to the cloud service but it still didn’t reappear, so i presumed was just a transient problem.
