No one said this behavior isn’t normal. I bet every oh install it there is polling once per minute.
What we have said is this behavior is not necessary. Not necessary is not the same as not normal.
And I’m just explaining the behavior you are seeing so that we can converge on a solution that solves this unnecessary polling for everyone, not just you.
No one is demanding you to set anything up right now. We are all busy and respect that.
No one said this behavior isn’t normal. . I bet every oh install it there is polling once per minute.
A) You said yourself so much polling is a mystery. When you say something is a mystery I assume you also consider it not normal.
For me it’s not because OH does this out of the box, that this is normal behavior.
You might called it expected OH behavior, yet not normal.
No software should do so much polling.
Especially when nothing needs to happen.
And I’m just explaining the behavior you are seeing
You already did that before.
know that, yet like I told before it will be later. No need to repeat that it needs to be done.
When I say it’s related I’m saying that probably somehow every minute the config file is checked to install new bindings. (except if it does that only when it’ can’t install a biding, yet thet fact that it checks openhab.jfrog.io every minute, I think this also happens every minute, so even that feels like overhead.
(that is using server resources)
Perhaps we have different definitions of normal. Usually normal means expected. It is perfectly valid for normal behavior to be unwanted behavior.
The polling is, as far as I can tell, the default behavior of maven In most configurations that use Maven like this the polling is desired because a new version of a package would trigger a new build. OH isn’t using maven in that way but no one noticed that this was the default behavior.
Will this be changed in a snapshot update soon?
Not until someone:
I don’t have the setup right not to track whether this polling is taking place on my system or I would have done so already. Yves understandably has other priorities right now so he can’t do the above. If you have monitoring in place it would help the process along if you would run this experiment to make sure that it works. If you verify it works (i.e. the polling stops but you can still install new bindings) I’ll even file the issue and make the PR if you don’t want to.
I changed the version.properties file to this:
Is this the right change?
openHAB Distribution Version Information
----------------------------------------
build-no : Build #1292
online-repo : https://openhab.jfrog.io/openhab/online-repo-snapshot/2.4@id=openhab@snapshots@update=never
Repository Version
----------------------------------------
openhab-distro : 2.4.0-SNAPSHOT
smarthome : 0.10.0-SNAPSHOT
openhab-core : 2.4.0-SNAPSHOT
openhab1-addons : 1.13.0-SNAPSHOT
openhab2-addons : 2.4.0-SNAPSHOT
karaf : 4.1.5
I still get connections to jfrog with this settings. I changed the file and made a service restart of openhab.
Any other things i could try?
EDIT:
I edited this two files. Still many connections to jfrog.
userdata/etc/org.ops4j.pax.url.mvn.cfg
userdata/etc/version.properties
No.
This is the right one.
Humph.
Maybe you could open an issue for this?
You know better where to post this and which information can be useful to solve this problem.
It wont do much good until we have a solution. It’s a pretty low level problem. It really isn’t a problem as much as it is an annoyance.
Sorry I did not have time to try out stuff, kids and exams are a higher priority.
It really isn’t a problem as much as it is an annoyance.
Well one connection every minute is still +10.000 unneeded connections every week for every installation.
That is not good for my network
yet that is terrible for jfrog.io
520.000 unneeded connection per openhab install, per year. is a pretty high number if you ask me.
I do agree. It should be fixed to be a good citizen for jfrog, though they haven’t complained so it apparently isn’t a problem for them. It really should be fixed. But if you compared to pretty much everything here listed as a bug is probably going to be a higher priority.
So far, you and Halloween are the only two to notice and complain about this problem. I’d like to look into it myself but like you have lots of other priorities taking away my time, and I don’t have the network set up to monitor the outgoing requests right now and doing so would take some time.
I found org.ops4j.pax.url.mvn.cfg
under /var/lib/openhab2/etc
when I change it
The line is replaced by org.ops4j.pax.url.mvn.repositories = https://openhab.jfrog.io/openhab/online-repo-snapshot/2.3@id=openhab@snapshots@update=never,https://openhab.jfrog.io/openhab/online-repo-snapshot/2.3@id=openhab@snapshots
so I guess something else has to be done. (or this is set hardcoded in openhab code…)
more: after I added that, my trick with remove = false no longer worked.
The line I added in the config stayed with remote false
when I remove the extra words, and then put remote false, then this line changes to
org.ops4j.pax.url.mvn.repositories =
So…I’m not alone 
after installing PiHole I also see this connections. Did you already find a solution for it?
Could the request that’s made every minute be caused by the job that checks the installed features every minute @Kai?
It was introduced in 2.2.0:
Sounds likely.
can this be turned of in any way?
Is there any progress? The problem still exists.