java delete all items in dynamodb

Question

Im trying to delete all items in my table in dynamodb but it does not work.

    try {
        ScanRequest scanRequest = new ScanRequest().withTableName(table);
        ScanResult scanResult = null;

        do {
            if (Check.nonNull(scanResult)) {
                scanRequest.setExclusiveStartKey(scanResult.getLastEvaluatedKey());
            }

            scanResult = client.scan(scanRequest);

            scanResult.getItems().forEach((Item) -> {
                String n1 = Item.get("n1").toString();
                String n2 = tem.get("n2").toString();
                DeleteItemSpec spec = new DeleteItemSpec().withPrimaryKey("n1", n1, "n2", n2);
                dynamodb.getTable(table).deleteItem(spec);
            });
        } while (Check.nonNull(scanResult.getLastEvaluatedKey()));
    } catch (Exception e) {
        throw new BadRequestException(e);
    }

n1 is my Primary partition key

n2 is my Primary sort key

Answer 1

PREAMBLE: While a scan operation is expensive, I was needing this answer for initialising a table for a test scenario (low volume). The table was being created by another process and I needed the test scenario on that table, I could therefore not delete and recreate the table.

ANSWER: given:

• DynamoDbClient db

• static String TABLE_NAME

• static String HASH_KEY

• static String SORT_KEY

   ScanIterable scanIterable = db.scanPaginator(ScanRequest.builder()
           .tableName(TABLE_NAME)
           .build());
   for(ScanResponse scanResponse:scanIterable){
       for( Map<String, AttributeValue> item: scanResponse.items()){
           Map<String,AttributeValue> deleteKey = new HashMap<>();
           deleteKey.put(HASH_KEY,item.get(HASH_KEY));
           deleteKey.put(SORT_KEY,item.get(SORT_KEY));
           db.deleteItem(DeleteItemRequest.builder()
                   .tableName(TRANSACTION_TABLE_NAME)
                   .key(deleteKey).build());
       }
   }
  

Answer 2

The best approach to delete all the items from DynamoDB is to drop the table and recreate it.

Otherwise, there are lot of read capacity and write capacity units being used which will cost you.

Dropping and recreating the table is the best approach.