OCP V4.2.1

V4.2.1 BP1

Version information

• Version: V4.2.1 BP1

• Previous version: V4.2.1

• Release date: December 08, 2023

Bug fixes

• Fixed the performance issue where resources of the sys tenant were excessively consumed when the data volume was large.

• Fixed the issue where the frontend page would become unresponsive when clicking any time component.

• Fixed the dependency on the MySQL client when deploying a standalone centralized cluster.

• Optimized the retrieval speed of alert information on the Backup and Recovery page of a cluster.

• Optimized the overall performance of OCP. Reduced the CPU and memory resource consumption required by OCP and significantly improved the response speed of web pages.

V4.2.1

OCP V4.2.1 officially supports OceanBase V4.2.1, enabling monitoring and alerting for end-to-end components of OceanBase, including support for custom monitoring, enhanced support for tenant- and session-level end-to-end tracing, and added monitoring and alerting capabilities for OBLB & OBDNS. For backup and recovery, it now supports Tencent Cloud COS, further improving support for storage media from major domestic cloud providers. Additionally, this version introduces 17 optimizations in cluster management, tenant management, monitoring and alerting, and task management, focusing on enhancing product usability to create a simpler and more user-friendly database management platform.

Version information

• Version: V4.2.1

• Previous version: V4.2.0

• Release date: November 21, 2023

• Supported upgrade path:

  • Currently, only upgrades from OCP V3.2.4 and later versions to the current version are supported.

  • For OCP V2.3.x and later, but earlier than V3.2.4, you must first upgrade to OCP V3.3.4.

  • For OCP versions earlier than V2.3.0, you must first upgrade to OCP V2.3.x, then to OCP V3.3.4.

Supported OceanBase Database versions

OCP V4.2.0 supports the following OceanBase Database versions:

• OceanBase V1.4.x

• OceanBase V2.1.x

• OceanBase V2.2.x

• OceanBase V3.1.x

• OceanBase V3.2.x

• OceanBase V4.x

New product features

OceanBase V4.x adaptation

• Supports full-link diagnostic capabilities at the tenant and session levels, helping users analyze transaction-level execution times from three perspectives: client > OBProxy > OceanBase cluster. This enables rapid root cause identification and timely fault resolution. This feature requires deploying the OpenSearch service and imposes version constraints on all components along the entire link. For details, see Query link.

• The Backup and Recovery module now supports integration with Tencent Cloud COS storage types (OceanBase Database must be V4.2.1 or later).

• Adapted to the __all_virtual_dump_tenant_info view to enable monitoring and alerting for tenant queue usage.

• Adapted to the semantics of the tenant=all parameter. This parameter takes effect only on user tenants and does not apply to the sys or meta tenant.

• Adapted to changes in the RS job view, adding task types such as ALTER_TENANT_PRIMARY_ZONE, ALTER_TENANT_LOCALITY, and ALTER_RESOURCE_TENANT_UNIT_NUM.

• Supports displaying OceanBase log disk usage information.

Feature enhancements

• OBProxy Management

  • Supports hot restart for OBProxy versions V3.2.8 and later, resolving the issue where maintenance or upgrade operations would impact online services.

  • Supports enabling, restarting, stopping OBProxy clusters, as well as batch starting, batch restarting (including hot restarts), and batch stopping of OBProxy nodes. This enhances OBProxy management capabilities and improves operational efficiency.

• Monitoring and Alerts

  • Supports user-defined monitoring capabilities. By defining collection items and monitoring items, and associating monitoring items with charts and alerts, users can create customized monitoring and alerting solutions. This addresses scenarios where OCP's built-in alerts cannot meet customer business requirements (currently only supporting alerts for related metrics within the sys tenant).

  • Supports alert capabilities for OBLB & OBDNS, offering a total of 16 alert metrics such as process status and process connectivity, thereby realizing alert capabilities across the entire OceanBase architecture.

  • Supports monitoring the active MemStore usage ratio. The view path is Tenant > Performance Monitoring > Performance and SQL > MEMStore Usage Percentage.

  • Supports monitoring the number of minor compactions per tenant, helping users prevent business impact caused by cluster-level major compactions triggered when the number of minor compactions in a single tenant of OceanBase V3.X exceeds the threshold.

  • Supports an alert when the total space usage of the OceanBase recycle bin exceeds 100 GiB, reducing the impact on customer business due to excessive partitions in the OceanBase V3.X recycle bin or untimely cleanup.

  • Supports alerts for inconsistencies between the OCP agent and the OCP version, preventing abnormal behavior during subsequent operations and enhancing system stability.

• Inspection

  • Supports connectivity detection between servers, including SQL ports and RPC ports, assisting users in quickly locating service interruptions caused by network or host failures.

  • Supports checking the usage rate of OceanBase macroblocks, helping users identify the actual data volume and disk usage of tenants and tables, thereby reducing disk fragmentation through major compaction operations.

Feature Optimizations

• For objects such as clusters, tenants, and OBProxy, or for hosts that are unavailable or offline, supports analysis of the cause of the exception, providing error information, error causes, and solutions to help users resolve faults and ensure stable business operation.

• Cluster Management:

  • Supports specifying startup parameters, such as memory_limit and system_memory, when adding or replacing OBServer nodes. This resolves issues like expansion failure and resource waste when hardware resources of machines within the cluster are inconsistent.

  • Supports specifying startup parameters, such as cpu_count and devname, when creating an arbitration service.

  • Added the option to perform a MemStore minor compaction during cluster restart, zone restart, OBServer restart, or OBServer process stop. This primarily addresses the issue of slow OBServer recovery without having performed a minor compaction.

  • Supports verifying the paths and sizes of the cluster log disk and data disk, such as detecting log disk size and whether the data disk and log disk are deployed on the same physical disk, and provides corresponding prompts. This avoids subsequent database runtime exceptions and enhances system stability.

  • Added a prompt for deadlock detection: This feature is not recommended for production systems running OceanBase V3.X.

  • Provides detection for support of Large System Extensions (LSE) instructions lower than ARM8.1, preventing deployment failures of OceanBase clusters.

• Host Management:

  • Supports associating a host IP with its corresponding OceanBase cluster or OBProxy cluster, and supports quick navigation to the respective cluster.

  • Supports quick navigation to the corresponding task for hosts in an operational state (task types include Delete Host and Reinstall OCP Agent).

• Monitoring and Alert Management:

  • Supports distinguishing between active alerts, historical alerts, and adds display reminders for active alerts in the status bar, enhancing users' attention to current active alerts.

  • Adds alert category options in Alerts > Alert Events, helping users quickly identify alert types and improving product usability.

  • Provides the affected host IP or tenant name within alert information (such as cluster alerts, tenant alerts, OBServer alerts), helping users quickly identify the scope of fault impact.

  • Changed the cluster name format to Cluster Name:Cluster ID and added primary/secondary cluster markers.

• Backup and Recovery Management:

  Changed the cluster name format to Cluster Name:Cluster ID and added primary/secondary cluster markers.

• Log Service Management:

  Added conditions such as sql_trace_id, trans_id, sess_id, and client_info in Link Query > Search Content, helping to correlate user links and effectively improve query efficiency.

• Task Management:

  • Supports displaying exception information for failed tasks and locating it to the position of the first error log, helping users efficiently pinpoint the cause of the fault.

  • Supports batch retrying of failed tasks. For scenarios with a large number of failures, such as Restart or ReInstall OCP Agent, this can significantly improve user O&M efficiency.

  • Supports quickly associating objects under O&M, such as clusters, tenants, OBProxies, and hosts, with tasks. This facilitates users quickly checking task status for subsequent O&M operations.

Product Behavior Changes

• Cluster Management: Deadlock Analysis has been changed to View Deadlock History Information.

• Tenant Management: Tenant (MySQL) > User Management > User List no longer supports displaying non-compliant usernames.

• Monitoring and Alert Management: The ability to upload configuration files has been removed from Alert Channel > Channel Type (Custom Script) to avoid issues with such file loss during upgrade scenarios.

Key Feature Interpretation

End-to-end Diagnostics at Tenant & Session Levels

OceanBase is a distributed database with extremely complex front-end to back-end call chains. When business experiences timeout or performance anomalies, users often cannot determine whether the issue lies with internal OceanBase components or network problems, and can only analyze based on experience and OBServer logs. This results in lengthy problem analysis and resolution times, affecting business stability. Therefore, OCP provides an end-to-end diagnostic mechanism to monitor the duration of the entire Client > OBProxy > OBServer chain, helping users quickly locate the root cause of problems.

If you need to view related log information in OCP, please first set the data mid-end module, link query collection, and OpenSearch (ES) related parameters. If only certain tenants or sessions need to be enabled, please disable the end-to-end diagnostic feature for other tenants, then go to the Log Service > Link Query page, and filter by time, tenant, search content (such as sql_id, trans_id, sess_id, etc.), and link duration. The current end-to-end diagnostic tracing is dimensioned by database transactions, and it collects data such as link duration, driver duration, OBProxy duration, and OB duration. Through simple duration analysis, you can identify the main current issues, and enter the corresponding trace to view specific durations (Span/Tag), thereby quickly identifying the real problem points.

Custom Monitoring

Custom monitoring allows you to customize monitoring charts and alerts based on your business scenarios when the built-in monitoring in OCP cannot meet your requirements. This helps you promptly perceive internal system changes and make timely and quick decisions. The entire feature module is divided into three tabs: Collection Items, Metrics, and Charts. Specifically:

1. Collection Items represent the raw monitoring data collected from different data sources. The current system has built-in 50+ collection items.

2. Metrics represent calculated metrics generated based on collection items, such as maximum value, minimum value, or average value, among dozens of other algorithms. You can select or create custom expressions based on your usage of monitoring data. The current system has built-in 200+ monitoring items.

3. A chart can consist of one or more metrics, and supports selecting a chart placement location and previewing the chart. Additionally, users can implement monitoring capabilities for corresponding metrics through the custom alert feature based on these metrics.

Fixed issues

OCP V4.2.1 primarily fixes the following issues:

• Fixed the issue where non-admin users in OCP could not access alert details.
• Fixed the issue where a rollback of an OBProxy creation task was incomplete, leading to failed redeployment of OBProxy.
• Fixed the issue where the tenant variable max_allowed_packet could not be modified on the OCP side.
• Fixed the issue where tenant session count statistics were incorrect.
• Optimized the upgrade speed from OCP V3.3.x to V4.2.x, which slowed down when there was a large amount of monitoring data.
• Optimized the volume of data collected, improving collection performance and stability.
• Optimized the timeout logic for O&M SQL queries, now supporting settings and performing more stably.
• Optimized monagent logs to avoid generating excessive useless debug logs.

Optimized alert behavior when MetaDB is unavailable.

Known issues

Version usage limitations

Hardware requirements

OCP-Server can be installed on physical servers or run in Docker containers. OCP-Server supports multi-node high-availability deployment.

The minimum hardware requirements for an OCP-Server node are listed in the following table.

OCP-Agent consumes very few resources and has no specific hardware requirements for the installation node.

Operating system requirements

The operating system requirements for installing the OCP server (including OCP-Agent) are listed in the following table.

Client requirements

Users typically access OCP services through a web browser. The client requirements are as follows.

If you need to access OCP from a device running iOS, the version requirements are listed in the following table.

For the best user experience, we recommend using a display with a resolution greater than 1440 × 810.