Monitoring is a vital component of maintaining the availability and performance of your databases.
Data Management for VMware Tanzu collects health and metric data for each database. You can view this data and use it to track the resource consumption, performance, and activity of your databases.
Database Status
Data Management for VMware Tanzu uses the database Status to reflect the availability the database, and in some cases to identify an in-progress operation or a critical operation that failed. The Status of a database is also affected by alerts that Data Management for VMware Tanzu may trigger on the database.
The Status of a database may be one of the following values:
| Status | Description |
|---|---|
| BACKUP_IN_PROGRESS | A backup of the database is in progress. |
| CONTROL_PLANE_UPDATE_IN_PROGRESS | A control plane software update is in progress. |
| CONTROL_PLANE_UPDATE_FAILED | A control plane software update failed. |
| CRITICAL | The Service Engine has at least one CRITICAL-level alert and no LOST_CONNECTIVITY or FATAL alert. |
| DB_ENGINE_UPDATE_IN_PROGRESS | A Service Engine software update is in progress. |
| DB_ENGINE_UPDATE_FAILED | A Service Engine software update failed. |
| DELETED | The database has been deleted. |
| DELETING | Deletion of the database has been initiated. |
| ERROR | Creation of the database failed. |
| FATAL | The database has at least one FATAL-level alert and no LOST_CONNECTIVITY alerts. |
| INIT | A create operation has been initiated for the database. |
| LOST_CONNECTIVITY | The database has at least one LOST_CONNECTIVITY alert. |
| ONLINE | The database was created successfully, is operating, and has no outstanding alerts. |
| OS_UPDATE_IN_PROGRESS | An operating system software update is in progress. |
| OS_UPDATE_FAILED | An operating system software update failed. |
| POWEREDOFF | The database VM is powered off. |
| POWEREDON | The database VM is running, but awaiting a health check. |
| WARNING | The database has at least one WARNING-level alert and no LOST_CONNECTIVITY, FATAL, or CRITICAL alerts. |
Viewing Database Status
The Status of a database is an indicator of the overall health of the database. You can view database status using the Data Management for VMware Tanzu console or API.
Perform the following procedure to examine the status of a database:
Select Databases from the left navigation pane.
This action displays the Databases view, a table that lists the provisioned database databases.
Examine the databases listed in the table, identify the database of interest, and navigate to that table row.
Examine the Status.
Viewing Database Health
The health of a database reflects the status of the services running in the VM, the status of certain resources that it consumes, and its connectivity to internal and external components. Database health is directly related to alerts that Data Management for VMware Tanzu may have triggered for the database. Refer to Database Alerts for details on the types and severity of alerts that Data Management for VMware Tanzu may trigger for an database.
Perform the following procedure to view the health of a database:
Select Databases from the left navigation pane.
This action displays the Databases view, a table that lists the provisioned databases.
Examine the databases listed in the table, identify the database of interest, and navigate to that table row.
Click the database VM Name.
The database information Details tab displays.
Select the Monitoring tab.
This action displays monitoring data for the database.
View the Health Status information.
Select the drop down menu in the top right corner to change the time-series aggregation period.
Database Alerts
Data Management for VMware Tanzu triggers an alert on a database when it encounters connectivity or resource issues on the database. You monitor these alerts in the Database Alerts view.
The Database Alerts view displays the following information:
- The VM Name identifies the name of the database VM that has alerts.
- The Status column identifies the status of the database.
- The Critical column identifies the number of CRITICAL-level alerts associated with the database.
- The Warning column identifies the number of WARNING-level alerts associated with the database.
- The Environment column identifies the infrastructure on which the database is running.
- The Triggered Time identifies the time at which the most recent alert was triggered.
- The Owner column identifies the Data Management for VMware Tanzu user that owns the database.
About the Alert Levels
Data Management for VMware Tanzu triggers alerts of the following levels:
- OK
- ONLINE
- WARNING
- CRITICAL
- FATAL
- LOST_CONNECTIVITY
About the Alert Types
<Any new DB alerts added/modified/deleted ?>
Alerts that Data Management for VMware Tanzu may trigger on a database include the following:
| Alert Name | Threshold | → Status | Alert Level | Triggered When | Impact |
|---|---|---|---|---|---|
| LOST_CONNECTIVITY | N/A | Lost Connectivity | CRITICAL | The database is unreachable by Data Management for VMware Tanzu. | Data Management for VMware Tanzu cannot reach the database but the Database client can reach it through the Application network. |
| CPU_HEALTH_ALERT | 90% | Critical | CRITICAL | vCPU utilization has reached 90%. | If sufficient CPU bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. |
| CPU_HEALTH_ALERT | 70% | Warning | WARNING | vCPU utilization has reached 70%. | If sufficient CPU bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. |
| DATA_DISK_HEALTH_ALERT | 90% | Fatal | CRITICAL | The data disk has reached 90% capacity. | If sufficient Database disk bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. Operations with respect to SSL configuration, backup, read replica, clone, binary logs, automated updates, and certificate refresh of databases are affected. |
| DATA_DISK_HEALTH_ALERT | 70% | Warning | WARNING | The data disk has reached 70% capacity. | If sufficient Database disk bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. |
| DATABASE_BIN_LOG_ALERT | N/A | Critical | CRITICAL | The transactions logs on the database are not getting copied to local storage. | PITR and Read replica operations are affected. |
| DATABASE_BIN_LOG_CLOUD_SYNC_ALERT | N/A | Warning | WARNING | The transaction logs on local and cloud storage are not in sync. | Database bin logs are not copied to the cloud storage. |
| DATABASE_SERVICE_ALERT | N/A | Fatal | CRITICAL | The database database engine is down. | Database client cannot reach Database engine and operations with respect to scale configuration, backup, read replica, clone, bin logs, automated updates, and certificate refresh of databases are affected. |
| MAX_CONNECTIONS_ALERT | 90% | Critical | CRITICAL | The number of open connections to the database is approaching the maximum connections allowed. | No new database connections are allowed from the database client. This alert is only available for MySQL and PostgreSQL databases. |
| MAX_CONNECTIONS_ALERT | 80% | Warning | WARNING | The number of open connections to the database has reached 80% of the maximum connections allowed. | No impact. This alert is only available for MySQL and PostgreSQL databases. |
| METRICS_ALERT | N/A | Warning | WARNING | Data Management for VMware Tanzu cannot pull metrics from the database. | Database metrics are not tracked on the Provider console. |
| NTP_SYNC_ALERT | N/A | Critical | CRITICAL | The database NTP service is down or not in sync. | If the NTP Service is down for more than 15 minutes, transaction logs, monitoring metrics, and database updates are affected. |
| SYSTEM_DISK_HEALTH_ALERT | 90% | Critical | CRITICAL | The system disk has reached 90% capacity. | If sufficient system disk bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. |
| SYSTEM_DISK_HEALTH_ALERT | 70% | Warning | WARNING | The system disk has reached 70% capacity. | If sufficient system disk bandwidth is not available, all Database VM and DB engine operations can get impacted and can result in failure. |
| TELEGRAF_SERVICE_ALERT | N/A | Warning | WARNING | The Telgraf service is not responding. | Database metrics are not tracked on the Provider console. |
Clearing Alerts
In some cases, you can clear certain alerts by restarting the affected service on the database VM.
| Alert Name | Affected Database OS Service Name |
|---|---|
| NTP_SYNC_ALERT | systemd-timesyncd |
| TELEGRAF_SERVICE_ALERT | telegraf |
| METRICS_ALERT | telegraf |
| METRICS_ALERT1 | influxdb |
| DATABASE_SERVICE_ALERT | dbengine |
1 If a METRICS_ALERT is raised on all databases, restart the influxdb.service on the Agent VM.
To clear an alert:
SSH into the database or Agent VM.
Restart the affected service. For example:
user@servinstvm$ systemctl restart telegraf.service
Addressing Other Alerts
If the Agent VM triggers a DATABASE_BIN_LOG_ALERT on a database, verify: the connection to Local Storage, the data disk space, and the service Engine status.
Database Metrics
Data Management for VMware Tanzu collects metric data for each database. You can view this data and use it to track the resource consumption, performance, and activity of your databases.
A database is created with NORMAL or ENHANCED monitoring. The default monitoring for a database is ENHANCED monitoring. The metrics for which Data Management for VMware Tanzu collects data in ENHANCED monitoring mode include the NORMAL metrics, plus additional service-specific metrics.
Modifying the Level of Monitoring
You can change the level of monitoring of a database after provisioning it from NORMAL or ENHANCED and vice-versa.
To modify the level of monitoring of a database:
Navigate to the DB Metrics pane in the Monitoring tab.
Click Edit.
The Monitoring Policy dialog displays.
Select Monitoring Type as either Normal or Enhanced, and then click SAVE.
Click the refresh icon on the top right corner of the pane to verify the change in the level of monitoring.
NORMAL Monitoring for MySQL and PostgreSQL Databases
Data Management for VMware Tanzu displays the following DB Metrics when NORMAL monitoring is enabled for a MySQL and PostgreSQL database:
- System Uptime - The time since the service or database VM restarted.
- Mysql Uptime (MySQL)
- Max Connections - The connection limit to the database.
- Active Connections per Second (PostgreSQL)
- Thread Resource Utilization (MySQL)
- CPU Usage % - The ratio of used to allocated CPU.
- Memory Usage % - The ratio of used to allocated memroy.
- Disk Usage % - The ratio of used to allocated disk.
NORMAL Monitoring for Microsoft SQL Server Databases
Data Management for VMware Tanzu displays the following DB Metrics when NORMAL monitoring is enabled for a Microsoft SQL Server database:
- System Uptime - The time since the service or database VM restarted.
- Sqlserver Uptime - The time since the service or Sqlserver VM restarted.
- Open Connections - The number of connections across frquent intervals of 20 minutes.
- CPU Usage % - The ratio of used to allocated CPU.
- Memory Usage % - The ratio of used to allocated memroy.
- Disk Usage % - The ratio of used to allocated disk.
ENHANCED Monitoring for MySQL and PostgreSQL Databases
In addition to the PostgreSQL statatistics displayed for NORMAL monitoring, the PostgreSQL statistics displayed when ENHANCED monitoring is in effect include:
- Write Throughput
- Read Throughput
- Commits & Rollbacks
- Deadlocks & Conflicts
In addition to the MySQL statatistics displayed for NORMAL monitoring, the MySQL statistics displayed when ENHANCED monitoring is in effect include:
- Innodb Pool Size
- Queries & Questions
- Bytes Received & Sent
- Slow Queries per Second
- InnoDB Buffer Usage %
- InnoDB Reads & Writes
- Command Reads & Writes
ENHANCED Monitoring for Microsoft SQL Server Databases
In addition to the Microsoft SQL Server statatistics displayed for NORMAL monitoring, the Microsoft SQL Server statistics displayed when ENHANCED monitoring is in effect include:
- Lock timeouts, Lock Waits & Deadlocks
- Reads & Writes
- Transactions (per Sec)
- Disk Read & Write Latency
Viewing the Metrics
You view metric data for a database in the Databases view, database Monitoring tab, DB Metrics pane.
By default, Data Management for VMware Tanzu displays the last 3 hours of aggregated metric data. You can change this time period (calculated from current time) via a drop-down in the upper right corner of the DB Metrics pane.
