Capella Columnar Services Alert Reference

  • Capella Columnar
  • reference
    +
    This reference lists the alerts that Capella Columnar services can emit, the conditions in which they occur, and a description for each.

    Metric-Based Alerts

    Alerts caused by changes to the usage of Capella Columnar resources.

    The notification messages you receive include information about the potential cause of the alert, as well as the actions you can take to resolve it. For help resolving the alerts, contact Couchbase support.

    Capella Columnar delivers alert notifications by:

    • Displaying a message banner in the Capella Columnar UI.

    • Keeping a record of all current and past alerts in the activity logs.

    • Sending email to users who enable email notifications for their accounts.

    For more information about receiving alert notifications, see Receive Alerts for a Cluster.

    Display Name Description

    High Heap Memory Usage

    The Analytics Service is using high heap memory, and has used at least 95% of its available heap memory. High memory usage events can impact the throughput of your service.

    Invalid Link Credentials

    Link credentials failed more than 10 times in the last 5 minutes when connecting, creating, or editing a Columnar link. A large number of wrong failed credentials can indicate a security concern.

    Links can be Columnar links or Kafka links created to connect to data sources like MongoDB, MSK, or S3 that you configured when setting up Columnar.

    Link Connection Failure

    Remote link connection from the data source cluster to the Columnar cluster has failed and the link is in a disconnected state. If link connection fails, all data updates from the data source stop.

    A remote link can be any Confluent, MSK, S3, or Couchbase Server link created to connect to data sources like MongoDB, S3, or MySQL that you configured when setting up Columnar.

    Make sure that you’re using the correct credentials, URL, and certificate so that the remote link has the correct data source cluster details.

    HTTP Request Timeouts

    The Analytics Service is experiencing a large number of timeouts in HTTP requests. When the total number of HTTP requests is > 100, 20% of requests are timing out, indicating a connectivity issue when processing Analytics requests.

    HTTP Request Failure

    The Analytics Service is experiencing a large number of failed HTTP requests. When the total number of HTTP requests is > 100, 20% of requests are failing, indicating potential issues with either the requests or the system.

    Scan Wait Timeouts

    The scan consistency requests are timing out. Scan wait timeouts happen when the maximum time to wait for datasets to be updated before executing a query is exceeded.

    Make sure to provide the appropriate value for the scan_consistency and scan_wait parameters.

    Record Parse Failure

    The Analytics Service is failing to parse records from a link. This indicates an issue with the records being parsed, with the parsing system, or with the link itself.