我们基于事件将数据从应用程序推送到DynamoDB。我们编写了一个lambda函数,该函数映射到此DynamoDB表。我们在应用程序中更新的记录已成功发送到DyanamoDB。但是,lambda通常会减少记录的数量。
如果我们的应用程序中更新了3条记录,则所有3条记录都将推送到Dynamo,但是lambda有时只给出2条或1条。
我们在lambda函数中构建了一个csv文件,因此该csv文件的记录数量也减少了。
我可以共享一些代码段,但是我不知道这与Lambda配置或我们用来将数据推送到Dynamo的Dynamo API有什么关系吗?
以下是Lambda代码:
public class ProductLambda implements RequestHandler<DynamodbEvent, String> {
private static final String M_NUM = "number";
private static final String FT_TABLE = "table";
private AmazonS3 s3 =
AmazonS3ClientBuilder.standard().withRegion(Regions.US_EAST_1).build();
private static final String S3_BUCKET = "bucket";
private static final String S3_FOLDER = "folder";
public ProductLambda() {
}
@Override
public String handleRequest(DynamodbEvent event, Context context) {
String[] appcol = new String[] {"COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4"};
String[] fttcol = new String[] {"COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4"};
createFolder(S3_BUCKET, S3_FOLDER, s3);
List<DynamodbStreamRecord> records = event.getRecords();
try {
List<LinkedHashMap<String, AttributeValue>> changedRecords = new ArrayList<LinkedHashMap<String, AttributeValue>>();
String ddbTable = "";
for (DynamodbStreamRecord record : records) {
String modelNumber = "";
String ddbARN = record.getEventSourceARN();
if (ddbARN != null) {
ddbTable = ddbARN.split("/")[1];
}
if (record != null) {
Map<String, AttributeValue> changedRec = record.getDynamodb().getNewImage();
if (changedRec != null) {
if (ddbTable.equalsIgnoreCase("table1")) {
LinkedHashMap<String, AttributeValue> staticColumnRecord = new LinkedHashMap<String, AttributeValue>();
List<String> list1 = new ArrayList<String>(Arrays.asList(appcol));
for (String key : list1) {
AttributeValue value = changedRec.get(key);
if (value != null) {
staticColumnRecord.put(key, value);
} else {
staticColumnRecord.put(key, new AttributeValue().withS("null"));
}
}
changedRecords.add(staticColumnRecord);
}
if (ddbTable.equalsIgnoreCase("table2")) {
LinkedHashMap<String, AttributeValue> staticColumnRecord = new LinkedHashMap<String, AttributeValue>();
List<String> list2 = new ArrayList<String>(Arrays.asList(fttcol));
for (String key : list2) {
AttributeValue value = changedRec.get(key);
if(key.equals("ft_mod")){
modelNumber = value.getS();
}
if (value != null) {
staticColumnRecord.put(key, value);
} else {
staticColumnRecord.put(key, new AttributeValue().withS("null"));
}
}
staticColumnRecord = getModelDetails(staticColumnRecord,modelNumber);
changedRecords.add(staticColumnRecord);
}
}
}
}
context.getLogger().log("[DEBUG] - Total Records Changed: " + changedRecords.size());
buildAndUploadCSV(changedRecords, context, ddbTable);
} catch (AmazonServiceException e) {
context.getLogger().log("[ERROR] - AmazonServiceException " + e);
} catch (AmazonClientException e) {
context.getLogger().log("[ERROR] - AmazonClientException " + e);
} catch (IOException e) {
context.getLogger().log("[ERROR] - IOException " + e);
}
context.getLogger().log("[DEBUG] - Total Records are: " + records.size());
return "Total records are: " + records.size();
}
/*
*
* @param staticColumnRecord
* @param modelNumber
* @param context
* @return
*/
private LinkedHashMap<String, AttributeValue> getModelDetails(LinkedHashMap<String, AttributeValue> staticColumnRecord,
String modelNumber) {
String[] modcol = new String[] { "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4", "COL1", "COL2", "COL3", "COL4" };
if (!modelNumber.isEmpty()) {
AmazonDynamoDB dynamoClient = AmazonDynamoDBClientBuilder.standard().build();
GetItemRequest getItemRequest = new GetItemRequest().withTableName(FT_TABLE)
.addKeyEntry(M_NUM, new AttributeValue().withS(modelNumber));
Map<String, AttributeValue> responseItem = dynamoClient.getItem(getItemRequest).getItem();
List<String> sf_mode_col = new ArrayList<String>(Arrays.asList(modcol));
if (responseItem != null) {
for (String key : sf_mode_col) {
AttributeValue val = responseItem.get(key);
if (val != null) {
staticColumnRecord.put(key, val);
} else {
staticColumnRecord.put(key, new AttributeValue().withS("null"));
}
}
}
}
return staticColumnRecord;
}
/*
* Creates a folder in the S3 bucket.
*/
private static void createFolder(String bucketName, String folderName, AmazonS3 client) {
ObjectMetadata metadata = new ObjectMetadata();
metadata.setContentLength(0);
InputStream emptyContent = new ByteArrayInputStream(new byte[0]);
PutObjectRequest putObjectRequest = new PutObjectRequest(bucketName, folderName, emptyContent, metadata);
PutObjectResult result = client.putObject(putObjectRequest);
}
/*
* Build CSV file of changed data and upload to S3 bucket.
*/
private void buildAndUploadCSV(List<LinkedHashMap<String, AttributeValue>> changedRecords, Context context, String tableName)
throws IOException {
Calendar calendar = Calendar.getInstance();
SimpleDateFormat formatter = new SimpleDateFormat("yyyyMMddHHmmss");
String outputName = tableName + "_" + formatter.format(calendar.getTime()) + ".csv";
List<String> headers = changedRecords.stream().flatMap(map -> map.keySet().stream()).distinct()
.collect(Collectors.toList());
Path filePath = Paths.get("/tmp", outputName);
try (OutputStreamWriter bw = new OutputStreamWriter(new FileOutputStream(filePath.toFile()), StandardCharsets.UTF_8))
{
StringBuffer headerContent = new StringBuffer();
for (String string : headers) {
headerContent.append(string);
headerContent.append(",");
}
if (headerContent.length() > 0 && headerContent.charAt(headerContent.length() - 1) == ',')
headerContent.setLength( headerContent.length() - 1 );
StringBuffer strBuilder = new StringBuffer();
for (LinkedHashMap<String, AttributeValue> lmap : changedRecords) {
StringBuilder stringBuilder = new StringBuilder("");
String sep = "";
for (Entry<String, AttributeValue> string2 : lmap.entrySet()) {
String value = string2.getValue().getS();
stringBuilder.append(sep).append("\"").append(value).append("\"");
sep = ",";
}
if (!stringBuilder.toString().isEmpty()) {
strBuilder.append(stringBuilder).append(System.getProperty("line.separator"));
}
}
headerContent.append("\n");
headerContent.append(strBuilder);
ByteArrayInputStream contentsAsStream = new ByteArrayInputStream(headerContent.toString().getBytes());
ObjectMetadata md = new ObjectMetadata();
md.setContentLength((long)headerContent.toString().getBytes().length);
md.setContentType("text/csv");
PutObjectRequest request = new PutObjectRequest("bucket", "folder/" + outputName,
contentsAsStream, md);
s3.putObject(request);
bw.flush();
bw.close();
} catch (IOException e) {
context.getLogger().log("[ERROR] - IOException " + e);
}
}
}