tencent cloud

문서Tencent Cloud Observability Platform

TDSQL-C MySQL

Download
포커스 모드
폰트 크기
마지막 업데이트 시간: 2026-08-05 10:10:34
AI 번역
The monitoring events generated by the resource instances and products that customers purchase and use on the cloud include:
Event Chinese Name
Event Name
Event type
Subordinate Dimension
Recoverable or Not
Event Description
Handling Method and Suggestions
All events
All events
cynosdb_mysql:ErrorEvent:All
Instance/Cluster
None
A catch-all event that matches all TDSQL-C MySQL events.
Generally, no separate handling is required, as a catch-all subscription has been configured in the event bus. For specific event types, refer to the respective subcategories.
Cloud API operations (CloudAudit)
Cloud API actions (CloudAudit)
cynosdb_mysql:CloudEvent:ApiCall
Account/API call
None
Delivers OpenAPI call operation records via CloudAudit.
Access and analyze audit logs, and pay attention to high-risk API calls from abnormal accounts (such as deleting instances, changing passwords, and so on).
Unplanned high availability switchover
Unplanned high availability
cynosdb_mysql:ErrorEvent:HA
Cluster/Instance
Yes
The primary node fails, and the system automatically switches to the standby node.
1. View the switch reason.
2. Check application reconnection.
3. Verify master-slave data consistency.
4. Troubleshoot the root cause on the source host.
Console operations (CloudAudit)
Console operations (CloudAudit)
cynosdb_mysql:CloudEvent:ConsoleCall
Account/Console
None
Delivers console operation records via CloudAudit.
Access audit logs and use CAM to review sensitive operations.
ProxyNotAvailable
Database agent exception
cynosdb_mysql:ErrorEvent:ProxyUnHealthy
Database proxy/Instance
Yes
The database agent process or node is abnormal, affecting read/write splitting routing.
1. View the status of the proxy node.
2. Check the number of proxy connections.
3. Restart the proxy or submit a ticket if necessary.
Mini program operations (CloudAudit)
Mini program operations (CloudAudit)
cynosdb_mysql:CloudEvent:MiniProgramCall
Account/Mini program
None
Delivers mini program channel operation records via CloudAudit.
Integrate with the mini program authorization policy to detect abnormal remote control and mitigate losses promptly.
OutOfMemory
OOM
cynosdb_mysql:ErrorEvent:AlarmMemOverKill
Instance/Node
Yes
Instance memory reaches its limit, triggering the OOM Killer.
1. View slow SQL queries and large queries.
2. Evaluate the upgrade.
3. Optimize the Buffer Pool/temporary table settings.
InstanceReadOnlyDueToStorageFull
Instance read-only due to storage full
cynosdb_mysql:ErrorEvent:AlarmDiskOverQuota
Instance/Cluster
Yes
Disk utilization reaches 100%, and the instance automatically switches to read-only mode.
1. Scale out resources or clean up data urgently.
2. Troubleshoot the source of space growth.
3. Restore read-only mode and then verify write operations.
Version upgrade events
Version upgrade events
cynosdb_mysql:ErrorEvent:UpgradeVersion
Instance/Cluster
Yes
Database kernel or agent version upgrade
1. Upgrade during off-peak hours.
2. Verify feature performance after the upgrade.
3. Check for incompatible SQL.
Cross-AZ latency
Cross-AZ latency
cynosdb_mysql:ErrorEvent:CrossAzReplicationLag
Cross-AZ replication
Yes
Cross-AZ primary/secondary replication latency exceeds the threshold.
1. Troubleshoot write pressure on the primary database.
2. Evaluate cross-AZ bandwidth.
3. Adjust the alarm threshold.
ProxyNodeRemoval
The database agent mount node is removed.
cynosdb_mysql:ErrorEvent:ProxyNodeFaultEliminated
Database proxy/Instance
Yes
The agent detects an abnormal backend node and removes it from the routing.
1. Monitor whether recovery occurs within 5 minutes.
2. Continuously remove the status of nodes that require troubleshooting.
RdmaDowngrade
RdmaDowngrade
cynosdb_mysql:ErrorEvent:RdmaDowngrade
Node/Network
Yes
The high-performance RDMA network is downgraded to TCP.
1. Monitor performance metrics.
2. File a ticket to troubleshoot the RDMA link.
3. Evaluate the switch traffic.
Planned high availability switchover
Planned high availability
cynosdb_mysql:ErrorEvent:FailureAvoid
Cluster/Instance
Yes
Planned HA switchover (such as version upgrade, host maintenance, and other preventive operations)
1. Confirm service reconnection.
2. Check the switchover duration.
3. Monitor performance after the switchover.
AuditLowRisk
AuditLowRisk
cynosdb_mysql:ErrorEvent:AuditLowRisk
Audit Rule
None
Triggers a low-risk audit rule (such as simple SELECT and routine DML).
1. Access audit log storage.
2. Periodically review the source of exceptions.
3. Restrict permissions when necessary.
AuditMediumRisk
AuditMediumRisk
cynosdb_mysql:ErrorEvent:AuditMediumRisk
Audit Rule
None
Triggers a medium-risk audit rule (such as batch UPDATE and sensitive table access).
1. Notify Security/DBA for review.
2. Evaluate permission revocation.
3. Troubleshoot the source of SQL.
AuditHighRisk
AuditHighRisk
cynosdb_mysql:ErrorEvent:AuditHighRisk
Audit Rule
None
Triggers a high-risk audit rule (such as DROP TABLE, TRUNCATE, and privilege bypass).
1. Immediately send an alarm notification to Security/Compliance.
2. Correlate with CloudAudit to locate the account.
3. Respond to emergencies when necessary.
Cluster isolation
ClusterIsolated
cynosdb_mysql:ErrorEvent:ClusterIsolated
Cluster
Yes
The cluster is isolated due to overdue payment/violation, with services stopped but data retained.
1. Top up immediately or handle the violation.
2. Evaluate the data retention period.
3. Verify after the isolation is removed.
Cross-server migration of Serverless instances is successful.
ServerlessExpandNodeBySwitchSuccess
cynosdb_mysql:ErrorEvent:ServerlessExpandNodeBySwitchSuccess
Serverless instance
Yes
Successful cross-server migration of Serverless instances during scaling.
Generally, no handling is required. Pay attention to business latency during the migration.
Recovery of storage usage exceeding the limit
StorageUseOverRecover
cynosdb_mysql:ErrorEvent:StopStorageOver
Instance/Cluster
Yes (Recovery Event)
The recovery event for "storage usage exceeding the limit", where disk utilization has fallen below the threshold.
Used to automatically close alarms. It is recommended to be used as an alarm chain closure event.
Storage usage exceeding the limit
StorageUseOver
cynosdb_mysql:ErrorEvent:ReportStorageOver
Instance/Cluster
Yes
Disk utilization exceeds the configured threshold (but not reaching 100%).
1. Scale out in advance.
2. Archive cold data.
3. Adjust the alarm threshold.
Cross-AZ semi-sync downgrade
SemiSyncDowngrade
cynosdb_mysql:ErrorEvent:SemiSyncDowngrade
Cross-AZ replication
Yes
Cross-AZ semi-sync is downgraded to async, affecting data reliability.
1. Troubleshoot cross-AZ network.
2. Monitor the RPO.
3. Switch back to semi-synchronous replication after the fix.
Global database unplanned Ops event
global database HA
cynosdb_mysql:ErrorEvent:GdnSwitchStandby
Global Database (GDN)
Yes
Unplanned primary/secondary switchover in GDN
1. Monitor cross-region link quality.
2. Evaluate the RTO/RPO.
3. Submit a ticket and follow up if necessary.
Isolating an Instance
InstanceIsolated
cynosdb_mysql:ErrorEvent:InstanceIsolated
Instance.
Yes
The instance is isolated due to overdue payment/violation.
1. Top up immediately.
2. Verify after the isolation is removed.
3. Long-term overdue payments carry a risk of reclamation.
Instance Deletion
InstanceDeleted
cynosdb_mysql:ErrorEvent:InstanceDeleted
Instance.
None
The instance is deleted.
1. Troubleshoot the deletion source.
2. Items can be restored during the recycle bin retention period.
3. We recommend integrating with the approval workflow.
Cluster Deletion
ClusterDeleted
cynosdb_mysql:ErrorEvent:ClusterDeleted
Cluster
None
The cluster is deleted.
1. Troubleshoot the deletion source.
2. Items can be restored during the recycle bin retention period.
3. We recommend integrating with the approval workflow.
DeviceFault
DeviceFault
cynosdb_mysql:ErrorEvent:DeviceFaultEvent
Host/Physical machine
Yes
Underlying physical/virtual machines have hardware or resource risks.
1. Monitor business metrics.
2. Tencent Cloud automatically performs migration adjustments.
3. Submit a ticket and follow up.
Analytic engine sync link interruption
LibradbSyncBreak
cynosdb_mysql:ErrorEvent:LibradbSyncBreak
Analytic engine/Sync link
Yes
Analytic engine sync link interruption.
1. Check the synchronization link.
2. Assess the impact on the analysis task.
3. Contact Tencent Cloud for follow-up.
Analytics engine OOM
LibradbOom
cynosdb_mysql:ErrorEvent:LibradbOom
Analytics engine instance
Yes
Analytics engine node OOM, which may cause analysis tasks to fail.
1. Troubleshoot large queries/complex aggregations.
2. Evaluate the upgrade.
3. Split slow analysis tasks.
Analytics engine instance deletion
LibradbInstanceDeleted
cynosdb_mysql:ErrorEvent:LibradbInstanceDeleted
Analytics engine instance
None
Analytics engine instance is deleted.
1. Troubleshoot the deletion source.
2. Roll back analysis task dependencies.
Analytics engine instance isolation
LibradbInstanceIsolated
cynosdb_mysql:ErrorEvent:LibradbInstanceIsolated
Analytics engine instance
Yes
Analytics engine instance is isolated due to overdue payment.
1. Top up immediately.
2. Verify after the isolation is removed.
3. Monitor downstream reporting tasks.
Analytic engine disk usage exceeding the limit.
LibradbDiskUsageOverThreshold
cynosdb_mysql:ErrorEvent:LibradbDiskUsageOverThreshold
Analytics engine instance
Yes
Analytic engine disk utilization exceeding the threshold.
1. Clean up temporary data.
2. Disk Scale-Out
3. Adjust the data retention policy.
Analytics engine node failure
LibradbNodeError
cynosdb_mysql:ErrorEvent:LibradbNodeError
Analytics engine node
Yes
Analytics engine node failure
1. Submit a ticket immediately.
2. Assess task migration.
3. Monitor task SLAs.
Auto-extend storage failure
AutoExtendStorageError
cynosdb_mysql:ErrorEvent:AutoExtendStorageError
Instance/Cluster
Yes
System auto-extend storage failed (due to overdue payment, insufficient quota, and so on).
1. Check the account balance.
2. Check the instance quota.
3. File a ticket to troubleshoot the scale-out link.
Auto-extend storage success.
AutoExtendStorageSuccess
cynosdb_mysql:ErrorEvent:AutoExtendStorageSuccess
Instance/Cluster
Yes
(Recovery Event)
Auto-extend storage success.
Generally, no handling is required. Record it in the Ops log.
Restarting an Instance
InstanceRestart
cynosdb_mysql:ErrorEvent:NodeRestart
Instance.
Yes
Instance node restart (such as kernel upgrade, underlying fault recovery, and so on)
1. Confirm service reconnection.
2. Check performance before and after the restart.
3. Submit a ticket if it occurs frequently.
Instance cross-NUMA
Instance cross numa
cynosdb_mysql:ErrorEvent:NumaDowngrade
Instance/Node
Yes
Instance migrated from a bound NUMA node to another NUMA node.
1. Assess the impact on latency-sensitive services.
2. Submit a ticket to follow up on NUMA scheduling.
3. Migrate back to the original NUMA if necessary.
Payment Overdue Alarm
Arrears Warning
cynosdb_mysql:ErrorEvent:ArrearsWarning
Instance.
None
Payment Overdue Alarm
1. Top up immediately.
2. Assess backup usage.
3. Modify the automatic backup policy.







도움말 및 지원

문제 해결에 도움이 되었나요?

피드백