Skip to content

[alerter]bugfix: preserve group alert recovery events #4168

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

Open
hutiefang76 wants to merge 1 commit into
base: master
Choose a base branch
from
Open

[alerter]bugfix: preserve group alert recovery events #4168

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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: 17 additions & 0 deletions ...alerter/src/main/java/org/apache/hertzbeat/alert/notice/impl/DbAlertStoreHandlerImpl.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -127,9 +127,26 @@ public GroupAlert store(GroupAlert groupAlert) {
}
// Save alert group
groupAlert.setAlertFingerprints(alertFingerprints.stream().toList());
refreshGroupStatus(groupAlert);
GroupAlert savedGroupAlert = groupAlertDao.save(groupAlert);
savedGroupAlert.setAlerts(groupAlert.getAlerts());
return savedGroupAlert;
}
}

private void refreshGroupStatus(GroupAlert groupAlert) {
List<String> alertFingerprints = groupAlert.getAlertFingerprints();
if (alertFingerprints == null || alertFingerprints.isEmpty()) {
return;
}
List<SingleAlert> alerts = singleAlertDao.findSingleAlertsByFingerprintIn(alertFingerprints);
if (alerts == null) {
return;
}
boolean hasFiringAlert = alerts.stream()
.anyMatch(alert -> CommonConstants.ALERT_STATUS_FIRING.equals(alert.getStatus()));
groupAlert.setStatus(hasFiringAlert
? CommonConstants.ALERT_STATUS_FIRING
: CommonConstants.ALERT_STATUS_RESOLVED);
}
}
13 changes: 11 additions & 2 deletions hertzbeat-alerter/src/main/java/org/apache/hertzbeat/alert/reduce/AlarmGroupReduce.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -294,13 +294,17 @@ private void sendGroupAlert(GroupAlertCache cache) {
AlertGroupConverge ruleConfig = groupDefines.get(cache.getGroupDefineName());
long repeatInterval = ruleConfig.getRepeatInterval() != null
? ruleConfig.getRepeatInterval() * MS_PER_SECOND : DEFAULT_REPEAT_INTERVAL;
boolean hasResolvedAlert = hasResolvedAlert(cache.getAlertFingerprints().values());

// Skip if within repeat interval
if (cache.getLastRepeatTime() > 0
if (!hasResolvedAlert
&& cache.getLastRepeatTime() > 0
&& now - cache.getLastRepeatTime() < repeatInterval) {
return;
}
cache.setLastRepeatTime(now);
if (!hasResolvedAlert) {
cache.setLastRepeatTime(now);
}
}

GroupAlert groupAlert = GroupAlert.builder()
Expand Down Expand Up @@ -392,6 +396,11 @@ private String determineGroupStatus(Collection<SingleAlert> alerts) {
.anyMatch(alert -> CommonConstants.ALERT_STATUS_FIRING.equals(alert.getStatus()))
? CommonConstants.ALERT_STATUS_FIRING : CommonConstants.ALERT_STATUS_RESOLVED;
}

private boolean hasResolvedAlert(Collection<SingleAlert> alerts) {
return alerts.stream()
.anyMatch(alert -> CommonConstants.ALERT_STATUS_RESOLVED.equals(alert.getStatus()));
}

@Data
private static class GroupAlertCache {
Expand Down
46 changes: 46 additions & 0 deletions ...ter/src/test/java/org/apache/hertzbeat/alert/notice/impl/DbAlertStoreHandlerImplTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,11 +19,13 @@

import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.ArgumentMatchers.anyList;
import static org.mockito.Mockito.never;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.when;
import org.apache.hertzbeat.alert.dao.GroupAlertDao;
import org.apache.hertzbeat.alert.dao.SingleAlertDao;
import org.apache.hertzbeat.common.constants.CommonConstants;
import org.apache.hertzbeat.common.entity.alerter.GroupAlert;
import org.apache.hertzbeat.common.entity.alerter.SingleAlert;
import org.junit.jupiter.api.BeforeEach;
Expand Down Expand Up @@ -122,4 +124,48 @@ public void testStoreExistingAlert() {
assertEquals(1L, groupAlert.getId());
}

@Test
public void storeResolvedAlertKeepsGroupFiringWhenOtherGroupAlertsStillFire() {
String groupKey = "instance:host1";
String resolvedFingerprint = "cpu";
String firingFingerprint = "memory";
groupAlert.setGroupKey(groupKey);
groupAlert.setStatus(CommonConstants.ALERT_STATUS_RESOLVED);
singleAlert.setFingerprint(resolvedFingerprint);
singleAlert.setStatus(CommonConstants.ALERT_STATUS_RESOLVED);

GroupAlert existingGroup = new GroupAlert();
existingGroup.setId(1L);
existingGroup.setAlertFingerprints(List.of(resolvedFingerprint, firingFingerprint));
when(groupAlertDao.findByGroupKey(groupKey)).thenReturn(existingGroup);

SingleAlert previousFiringAlert = new SingleAlert();
previousFiringAlert.setId(1L);
previousFiringAlert.setFingerprint(resolvedFingerprint);
previousFiringAlert.setStatus(CommonConstants.ALERT_STATUS_FIRING);
previousFiringAlert.setStartAt(1000L);
previousFiringAlert.setActiveAt(2000L);
previousFiringAlert.setTriggerTimes(3);
when(singleAlertDao.findByFingerprint(resolvedFingerprint)).thenReturn(previousFiringAlert);

SingleAlert savedResolvedAlert = new SingleAlert();
savedResolvedAlert.setId(1L);
savedResolvedAlert.setFingerprint(resolvedFingerprint);
savedResolvedAlert.setStatus(CommonConstants.ALERT_STATUS_RESOLVED);
when(singleAlertDao.save(any(SingleAlert.class))).thenReturn(savedResolvedAlert);

SingleAlert existingFiringAlert = new SingleAlert();
existingFiringAlert.setFingerprint(firingFingerprint);
existingFiringAlert.setStatus(CommonConstants.ALERT_STATUS_FIRING);
when(singleAlertDao.findSingleAlertsByFingerprintIn(anyList()))
.thenReturn(List.of(savedResolvedAlert, existingFiringAlert));

when(groupAlertDao.save(any(GroupAlert.class))).thenAnswer(invocation -> invocation.getArgument(0));

GroupAlert savedGroupAlert = dbAlertStoreHandler.store(groupAlert);

assertEquals(CommonConstants.ALERT_STATUS_FIRING, savedGroupAlert.getStatus());
assertEquals(CommonConstants.ALERT_STATUS_FIRING, groupAlert.getStatus());
}

}
42 changes: 41 additions & 1 deletion hertzbeat-alerter/src/test/java/org/apache/hertzbeat/alert/reduce/AlarmGroupReduceTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.ArgumentMatchers.argThat;
import static org.mockito.Mockito.never;
import static org.mockito.Mockito.clearInvocations;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.when;
import java.util.Arrays;
Expand All @@ -55,6 +56,7 @@
import java.util.concurrent.atomic.AtomicBoolean;
import java.util.concurrent.atomic.AtomicInteger;
import org.apache.hertzbeat.alert.dao.AlertGroupConvergeDao;
import org.apache.hertzbeat.common.constants.CommonConstants;
import org.apache.hertzbeat.common.config.VirtualThreadProperties;
import org.apache.hertzbeat.common.entity.alerter.AlertGroupConverge;
import org.apache.hertzbeat.common.entity.alerter.SingleAlert;
Expand Down Expand Up @@ -83,7 +85,7 @@ void setUp() {
when(alertGroupConvergeDao.findAlertGroupConvergesByEnableIsTrue())
.thenReturn(Collections.emptyList());
alarmGroupReduce = new AlarmGroupReduce(alarmInhibitReduce, alertGroupConvergeDao,
new VirtualThreadProperties(), false);
new VirtualThreadProperties(false, null, null, null, null, null, null), false);
}

@AfterEach
Expand Down Expand Up @@ -129,6 +131,35 @@ void whenMatchingGroupRule_shouldGroup() {
verify(alarmInhibitReduce, never()).inhibitAlarm(any()); // Should not send immediately due to group wait
}

@Test
void resolvedAlertInFiringGroupShouldBypassRepeatThrottle() {
AlertGroupConverge rule = new AlertGroupConverge();
rule.setName("test-rule");
rule.setGroupLabels(Collections.singletonList("instance"));
rule.setGroupWait(0L);
rule.setGroupInterval(0L);
rule.setRepeatInterval(60L);
alarmGroupReduce.refreshGroupDefines(Collections.singletonList(rule));

alarmGroupReduce.processGroupAlert(createAlert("cpu", CommonConstants.ALERT_STATUS_FIRING));
alarmGroupReduce.dispatchCheckAndSendGroups();
verify(alarmInhibitReduce).inhibitAlarm(argThat(group ->
CommonConstants.ALERT_STATUS_FIRING.equals(group.getStatus())
&& group.getAlerts().size() == 1
&& group.getAlerts().get(0).getFingerprint().equals("cpu")));
clearInvocations(alarmInhibitReduce);

alarmGroupReduce.processGroupAlert(createAlert("cpu", CommonConstants.ALERT_STATUS_FIRING));
alarmGroupReduce.processGroupAlert(createAlert("memory", CommonConstants.ALERT_STATUS_RESOLVED));
alarmGroupReduce.dispatchCheckAndSendGroups();

verify(alarmInhibitReduce).inhibitAlarm(argThat(group ->
CommonConstants.ALERT_STATUS_FIRING.equals(group.getStatus())
&& group.getAlerts().stream()
.anyMatch(alert -> "memory".equals(alert.getFingerprint())
&& CommonConstants.ALERT_STATUS_RESOLVED.equals(alert.getStatus()))));
}

@Test
void dispatchCheckAndSendGroupsRunsOnVirtualThread() throws Exception {
CountDownLatch latch = new CountDownLatch(1);
Expand Down Expand Up @@ -173,6 +204,15 @@ private Map<String, String> createLabels(String... keyValues) {
return labels;
}

private SingleAlert createAlert(String fingerprint, String status) {
return SingleAlert.builder()
.fingerprint(fingerprint)
.status(status)
.labels(createLabels("instance", "host1"))
.annotations(Collections.emptyMap())
.build();
}

private static final class TestAlarmGroupReduce extends AlarmGroupReduce {

private final CountDownLatch virtualThreadLatch;
Expand Down
Loading