Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[#1114] feat: introduce hdfs host as the total_hadoop_write_data metric label #1107

Merged
merged 2 commits into from
Aug 10, 2023
Merged

[#1114] feat: introduce hdfs host as the total_hadoop_write_data metric label #1107

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
45f0554
feat: introduce hdfs host as the total_hadoop_write_data metric label
zuston Aug 7, 2023
04134cd
move lines
zuston Aug 10, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 15 additions & 2 deletions server/src/main/java/org/apache/uniffle/server/ShuffleServerMetrics.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -85,6 +85,7 @@ public class ShuffleServerMetrics {
private static final String STORAGE_FAILED_WRITE_LOCAL = "storage_failed_write_local";
private static final String STORAGE_SUCCESS_WRITE_LOCAL = "storage_success_write_local";
private static final String STORAGE_HOST_LABEL = "storage_host";
public static final String STORAGE_HOST_LABEL_ALL = "ALL";
public static final String STORAGE_TOTAL_WRITE_REMOTE = "storage_total_write_remote";
public static final String STORAGE_RETRY_WRITE_REMOTE = "storage_retry_write_remote";
public static final String STORAGE_FAILED_WRITE_REMOTE = "storage_failed_write_remote";
Expand Down Expand Up @@ -125,7 +126,6 @@ public class ShuffleServerMetrics {
public static Counter.Child counterTotalReadTime;
public static Counter.Child counterTotalFailedWrittenEventNum;
public static Counter.Child counterTotalDroppedEventNum;
public static Counter.Child counterTotalHadoopWriteDataSize;
public static Counter.Child counterTotalLocalFileWriteDataSize;
public static Counter.Child counterTotalRequireBufferFailed;
public static Counter.Child counterTotalRequireBufferFailedForHugePartition;
Expand Down Expand Up @@ -158,10 +158,13 @@ public class ShuffleServerMetrics {
public static Gauge.Child gaugeEventQueueSize;
public static Gauge.Child gaugeAppNum;
public static Gauge.Child gaugeTotalPartitionNum;

public static Counter counterRemoteStorageTotalWrite;
public static Counter counterRemoteStorageRetryWrite;
public static Counter counterRemoteStorageFailedWrite;
public static Counter counterRemoteStorageSuccessWrite;
public static Counter counterTotalHadoopWriteDataSize;

private static String tags;
public static Counter counterLocalFileEventFlush;
public static Counter counterHadoopEventFlush;
Expand Down Expand Up @@ -231,6 +234,14 @@ public static void incStorageFailedCounter(String storageHost) {
}
}

public static void incHadoopStorageWriteDataSize(String storageHost, long size) {
if (StringUtils.isEmpty(storageHost)) {
return;
}
counterTotalHadoopWriteDataSize.labels(tags, storageHost).inc(size);
counterTotalHadoopWriteDataSize.labels(tags, STORAGE_HOST_LABEL_ALL).inc(size);
}

private static void setUpMetrics() {
counterTotalReceivedDataSize = metricsManager.addLabeledCounter(TOTAL_RECEIVED_DATA);
counterTotalWriteDataSize = metricsManager.addLabeledCounter(TOTAL_WRITE_DATA);
Expand All @@ -253,7 +264,9 @@ private static void setUpMetrics() {
counterTotalDroppedEventNum = metricsManager.addLabeledCounter(TOTAL_DROPPED_EVENT_NUM);
counterTotalFailedWrittenEventNum =
metricsManager.addLabeledCounter(TOTAL_FAILED_WRITTEN_EVENT_NUM);
counterTotalHadoopWriteDataSize = metricsManager.addLabeledCounter(TOTAL_HADOOP_WRITE_DATA);
counterTotalHadoopWriteDataSize =
metricsManager.addCounter(
TOTAL_HADOOP_WRITE_DATA, Constants.METRICS_TAG_LABEL_NAME, STORAGE_HOST_LABEL);
counterTotalLocalFileWriteDataSize =
metricsManager.addLabeledCounter(TOTAL_LOCALFILE_WRITE_DATA);
counterTotalRequireBufferFailed = metricsManager.addLabeledCounter(TOTAL_REQUIRE_BUFFER_FAILED);
Expand Down
7 changes: 6 additions & 1 deletion server/src/main/java/org/apache/uniffle/server/storage/HadoopStorageManager.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,12 @@ public class HadoopStorageManager extends SingleStorageManager {
@Override
public void updateWriteMetrics(ShuffleDataFlushEvent event, long writeTime) {
super.updateWriteMetrics(event, writeTime);
ShuffleServerMetrics.counterTotalHadoopWriteDataSize.inc(event.getSize());
Storage storage = event.getUnderStorage();
if (storage == null) {
LOG.warn("The storage owned by event: {} is null, this should not happen", event);
return;
}
ShuffleServerMetrics.incHadoopStorageWriteDataSize(storage.getStorageHost(), event.getSize());
}

@Override
Expand Down
36 changes: 36 additions & 0 deletions server/src/test/java/org/apache/uniffle/server/ShuffleServerMetricsTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,6 +131,42 @@ private void validateMetrics(
assertTrue(bingo);
}

@Test
public void testHadoopStorageWriteDataSize() {
// case1
String host1 = "hadoop-cluster01";
ShuffleServerMetrics.incHadoopStorageWriteDataSize(host1, 1000);
assertEquals(
1000.0,
ShuffleServerMetrics.counterTotalHadoopWriteDataSize
.labels(Constants.SHUFFLE_SERVER_VERSION, host1)
.get());

// case2
ShuffleServerMetrics.incHadoopStorageWriteDataSize(host1, 500);
assertEquals(
1500.0,
ShuffleServerMetrics.counterTotalHadoopWriteDataSize
.labels(Constants.SHUFFLE_SERVER_VERSION, host1)
.get());

// case3
String host2 = "hadoop-cluster2";
ShuffleServerMetrics.incHadoopStorageWriteDataSize(host2, 2000);
assertEquals(
2000.0,
ShuffleServerMetrics.counterTotalHadoopWriteDataSize
.labels(Constants.SHUFFLE_SERVER_VERSION, host2)
.get());

// case4
assertEquals(
3500.0,
ShuffleServerMetrics.counterTotalHadoopWriteDataSize
.labels(Constants.SHUFFLE_SERVER_VERSION, ShuffleServerMetrics.STORAGE_HOST_LABEL_ALL)
.get());
}

@Test
public void testStorageCounter() {
// test for local storage
Expand Down