Get early explicit errors, prevents distinct-errrors from being applied.

If not actually distinct, PowerQuery force early query failure.pq

/// if conditions fail, don't allow the query
// to silently complete as "successful"
let
    Source = Sql.Databases("server"),
    nin = Source{ [Name="instance"] }[Data],
    PingLog = nin{ [Schema="dbo",Item="PingLog"] }[Data],

    // optional function: prints list values as a `csv` without having to drill down
    list_to_string = (text_list as list) as text =>
        let 
            seperator = ", ",
            template = "{ #{0} }",
            combined_string = Text.Combine(text_list, seperator)
        in
            Text.Format(template, {combined_string}),

    // Sorted Rows was the final step before error handling
    final_table = #"Sorted Rows", 

    // column names used for testing whether a table is distinct 
    // ie: the Table's primary keys
    distinct_columns = { "DatetimeUtc", "Destination", "Source" }, 

    final_table_otherwise_fail = if 
        Table.IsDistinct(
            final_table,
            distinct_columns
        )
    then
        final_table as table
    else 
        error [
            Reason = "ColumnsNotDistinct",
            Message = "Required distinct columns found duplicate records",
            Detail = "Failed PKs = " & list_to_string( distinct_columns )
        ]

in 
    final_table_otherwise_fail