Unhandled Error in case of unknown clusterId in Resource Loader

[sentry-io]

From Sentry Issue: VSCODE-EXTENSION-HR

Error: Unknown connectionId b678c432-8763-4465-816a-ab9736f10b0f
  File "..\src\storage\resourceLoader.ts", line 60, in {_ResourceLoader#2}.getInstance
    throw new Error(`Unknown connectionId ${connectionId}`);
  File "..\src\viewProviders\topics.ts", line 205, in {_TopicViewProvider#2}.updateTreeViewDescription
    const loader = ResourceLoader.getInstance(cluster.connectionId);
  File "..\src\viewProviders\topics.ts", line 184, in currentKafkaClusterChangedSub
    await this.updateTreeViewDescription();
  File "app:///extensionHostProcess.js", line 30, in P.B
  File "app:///extensionHostProcess.js", line 30, in P.fire
...
(3 additional frame(s) were not displayed)

[noeldevelops]

In the stack trace it seems like the user changed the Current Cluster and the resource loader somehow got passed an unknown ID to the updateTreeViewDescription() function.

At minimum, let's handle the error to prevent crash/Sentry alert. But I'm curious how this happened and how could we reproduce?