[knx] Handle exceptions during initial read (openhab#12520)

fixes openhab#7239

Signed-off-by: Holger Friedrich <mail@holger-friedrich.de>
Signed-off-by: Nick Waterton <n.waterton@outlook.com>

bundles/org.openhab.binding.knx/src/main/java/org/openhab/binding/knx/internal/client/AbstractKNXClient.java

@@ -14,6 +14,7 @@
 
 import java.time.Duration;
 import java.util.Set;
+import java.util.concurrent.CancellationException;
 import java.util.concurrent.CopyOnWriteArraySet;
 import java.util.concurrent.LinkedBlockingQueue;
 import java.util.concurrent.ScheduledExecutorService;
@@ -291,6 +292,8 @@ private void readNextQueuedDatapoint() {
                 logger.trace("Sending a Group Read Request telegram for {}", datapoint.getDatapoint().getMainAddress());
                 processCommunicator.read(datapoint.getDatapoint());
             } catch (KNXException e) {
+                // Note: KnxException does not cover KnxRuntimeException and subclasses KnxSecureException,
+                // KnxIllegArgumentException
                 if (datapoint.getRetries() < datapoint.getLimit()) {
                     readDatapoints.add(datapoint);
                     logger.debug("Could not read value for datapoint {}: {}. Going to retry.",
@@ -299,9 +302,15 @@ private void readNextQueuedDatapoint() {
                     logger.warn("Giving up reading datapoint {}, the number of maximum retries ({}) is reached.",
                             datapoint.getDatapoint().getMainAddress(), datapoint.getLimit());
                 }
-            } catch (InterruptedException e) {
+            } catch (InterruptedException | CancellationException e) {
                 logger.debug("Interrupted sending KNX read request");
                 return;
+            } catch (Exception e) {
+                // Any other exception: Fail gracefully, i.e. notify user and continue reading next DP.
+                // Not catching this would end the scheduled read for all DPs in case of an error.
+                // Severity is warning as this is likely caused by a configuration error.
+                logger.warn("Error reading datapoint {}: {}", datapoint.getDatapoint().getMainAddress(),
+                        e.getMessage());
             }
         }
     }