Added more details to troubleshooting around restore errors and warnings

docs/TROUBLESHOOTING.md

@@ -99,6 +99,18 @@ This section includes how to debug a failed restore. For more specific issues re
 
 <hr style="height:1px;border:none;color:#333;">
 
+<h3 align="center">Restore Errors and Warnings<a id="creds"></a></h3>
+
+Restore errors and warnings are shown in the `veleror describe` output in three groups:
+1. Velero: Warnings or errors that relate to the overall restore workflow but not to a particular resource will be shown here.
+2. Cluster: Warnings or errors that relate to particular cluster-scoped resources will be shown here
+3. Namespaces: Warnings or errors that relate to particular namespaced resources will be shown here, grouped by namespace
+
+One or more errors in any of the above categories will cause a Restore to be `PartiallyFailed` rather than `Completed`. Warnings will not cause a change to the completion status.
+
+Note that if there are "Velero" errors (but no resource-specific errors), it is possible that the restore completed without any actual problems with restoring workloads, but careful validation of post-restore applications is advisable.
+For example, if there are PodVolumeRestore and/or Node Agent-related errors, check the status of PodVolumeBackups and DataDownloads -- if none of these are failed or still running, then volume data may have been fully restored.
+
 <h1 align="center">Common Issues and Misconfigurations<a id="misconfig"></a></h1>
 
 <h3 align="center">Credentials Secret Not Properly Formatted<a id="creds"></a></h3>