Yes, this is a known problem at the moment. There is a NodeCache in the background, it is just configured by default with a short timeout, to ensure that the nodes are always up to date.

You can try to increase the timeout, for example with

or with an even bigger value (in milliseconds).

The effect is that the nodes are no longer read again and again from the server. Just note that the Value is also not refreshed any more. The Value refresh is the reason why we currently define the timeout to be just 1000 ms by default; so we can “guarantee” that the latest value is available from the node, but it also means that the complete nodes are re-read all the time.

We will need to figure out a way to keep the nodes in the cache and still read the value from the server whenever it needs to be, to guarantee that it is up to date.

Reading several nodes with a single read – or enabling some kind of lazy loading would be good ideas for future development as well.

Yes, I believe you: it fetches currently every node separately and also requests the references and properties separately for each node. So it is making a lot of calls, which could theoretically be combined. In practice, it is more complicated, but we will need to put effort in this in future.