Openhab2 / Zwave / Textual Configuration: A Discussion

First of all I consider OpenHAB2 to be a gorgeous product. Nevertheless I am currently thinking about a “larger scale” installation in a family house, leading to wealth of sensors and actuators (around 100+). For this reason the possibility for …

A) a fast disaster recovery and …
B) a smooth transition to newer OpenHAB-versions

is important for me.

A) is mitigated by using an backupable Z-Stick (Gen 5) in case of hardware failure. This allows me to continue with a new stick by restoring the Z-Stick configuration without including all devices manually. Further I try to use the textual configuration for things, items, persistence, sitemaps which allows me a fast recovery or a smooth transition to a new OpenHAB-installation (B).

With this in mind I discovered a few shortcomings (or features) which I want to place here for discussion:

  1. When a thing is defined textually, it is also automatically discovered which crowds up the inbox and may lead to duplicates for the same thing: Would it be of advantage to hide textually defined things from the inbox, which means to blacklist those node_id(s) mentioned in the textual configuration?
  2. If the Z-Stick is defined textually as a bridge, the include/exclude commands are not working any more: Would it be of advantage to allow include/exclude of ZWave-Devices even if the Z-Stick bridge is defined textually? (Currently I use a second instance of OpenHAB to do a secure inclusion)
  3. If a thing is defined textually, it is not possible to change the “configuration parameters” in HABmin. For example if I want to calibrate a temperature sensor, I have to switch to a non-textual configuration to do this: Is this restriction necessary? Would it be possible to set those “configuration parameters” in HABmin? Would it be nice, if one could set those configuration properties in the textual configuration?
  4. What is the purpose of the xml-files in userdata/zwave? I discovered, when I do an inclusion in a non textually defined environment a file with network_xxxxxxxx__node_y.xml is created. If I transfer the Z-Stick to another instance a similar file is created, but it differs. It is mostly striking, when a secure device is used: The secure classes are not enumerated in the xml, but secure sensor readings are properly received.
    What is the exact purpose of the xml-files?
    Would it be possible to query all those parameters from the Z-Stick or the Z-wave network and use those xmls as cache only (if required)?

Best regards,

brstein
1 Like

There are some updates going into ESH that should allow this to be resolved in the near future.

ESH doesn’t allow configuration of things that are defined in text files. There has been discussion about adding something to allow this, but so far, there’s nothing that allows this. Since exclude is implemented as a configuration parameter, this also doesn’t work.

See above - same issue.

They are used to persist the state between restarts. It speeds up the binding startup - if the files aren’t there, the binding will be slower to start, but it won’t hurt.

1 Like