In August 2026, GreenNode introduces major product updates across AI Stack, cloud infrastructure, backup, monitoring, security, and managed databases. The central theme is clear: help engineering teams operate infrastructure and AI agents faster through natural-language workflows, command-line automation, and stronger operational visibility.

This month’s highlights include GreenNode MCP for vServer and VBackup, an MCP Server for VKS, AgentBase CLI, expanded AI agent integrations, a Grafana datasource for vMonitor, CNAME-based DNS configuration for vWAF, and PostgreSQL Cluster availability in the Hanoi region.

Highlights at a glance

  • Use GreenNode MCP to manage vServer and VBackup through compatible AI assistants and custom AI agents.
  • Connect AgentBase agents to Microsoft 365, GitHub, Jira, Confluence, Monday.com, and other work tools through MCP Connectors.
  • Manage and deploy AI agents, Kubernetes, and virtual infrastructure from the terminal with CLI workflows.
  • Monitor vMonitor metrics and logs from Grafana dashboards and alerting workflows.
  • Gain clearer backup scheduling information and email alerts for missed backup executions.
  • Configure vWAF DNS with CNAME records in addition to A records.
  • Deploy PostgreSQL Cluster workloads in the Hanoi region with multi-node architecture, automatic failover, and separate read/write endpoints.

AI Stack: Bring agents into daily workflows

MCP Connectors for AgentBase

AI agents built on AgentBase can now connect directly to widely used business and engineering tools through MCP Connectors. Available integrations include Microsoft 365 services such as Teams, Outlook, and SharePoint; GitHub; Atlassian products including Jira and Confluence; and Monday.com. Teams can also request additional integrations based on their operational needs.

This reduces the work required to connect an AI agent with the systems where business context already lives. Instead of building a custom integration layer for each tool, teams can equip agents to retrieve context and support workflows across collaboration, project management, documentation, source control, and enterprise productivity platforms.

For example, an internal engineering-support agent could use GitHub to inspect repository context, Jira to review ticket status, Confluence to reference technical documentation, and Microsoft Teams to share a summarized response with the relevant team. The value is not simply access to more systems—it is the ability to make an agent useful within the tools employees already use each day.

AgentBase CLI and MCP support

AgentBase now supports command-line management and deployment of AI agents. This gives developer and platform teams a more automation-friendly option than relying only on the Portal interface. Agent lifecycle tasks can be incorporated into existing scripts, development workflows, and CI/CD pipelines.

In addition, AgentBase can connect directly to AI coding tools through the Model Context Protocol (MCP). This is designed for teams that want to work with agents from environments such as AI-enabled developer tools, rather than switching between separate management interfaces.

Security Runtime for enterprise agent access control

AgentBase Runtime adds controls for restricting and authenticating inbound access. When creating a Runtime, users can define allowed IP ranges and choose an inbound authentication method:

  • IAM Permissions for identity- and permission-based access control.
  • JWT for token-based request authentication.
  • No auth for use cases that do not require an authentication layer.

These controls help teams apply security requirements at the Runtime layer, particularly when exposing agent endpoints to internal systems, applications, or selected external clients. IP allowlisting limits where requests can originate, while authentication settings determine how callers prove they are authorized to access the Runtime.

Clearer MaaS model catalog

GreenNode Model-as-a-Service catalog is now organized into two clearly labeled groups:

  • Self-host: Models operated by GreenNode.
  • Partner: Models provided by third-party partners.

The updated catalog also includes improved pricing for multiple models. This clearer model classification can help teams make more informed choices based on factors such as operational ownership, workload requirements, pricing, and preferred model providers. Existing legacy models were set to renew automatically until September 2, 2026.

Kubernetes and cloud operations with MCP

VKS MCP Server: Manage Kubernetes in natural language

GreenNode Kubernetes Service (VKS) now supports an MCP Server that connects Kubernetes environments with compatible AI assistants, including Claude, Cursor, and VS Code-based workflows. Users can manage clusters, node groups, and Kubernetes resources using natural-language instructions instead of completing every operation manually in the Portal.

This can be particularly useful for developers and platform engineers who already work inside AI-assisted coding environments. Rather than interrupting their workflow to navigate multiple cloud-console pages, they can use an MCP-connected assistant to perform supported operational tasks with relevant Kubernetes context.

A practical use case could be asking an AI assistant to review cluster resources, inspect a Node Group configuration, or help manage a Kubernetes resource while the engineer remains in their existing development environment. Teams should still apply established review and access-control practices for production operations.

GreenNode MCP for vServer

GreenNode MCP brings AI-assisted infrastructure management to vServer. With more than 183 integrated tools, it enables users to manage key vServer resources through natural-language requests, including:

  • Creating, starting, stopping, and managing instances.
  • Configuring network resources.
  • Managing volumes and snapshots.
  • Working with security groups and SSH keys.

GreenNode MCP is read-only by default. Write access must be explicitly granted, helping teams separate infrastructure discovery from actions that modify running environments. The integration supports Claude Desktop, Cursor IDE, and custom AI agents.

This default-read-only approach is important for infrastructure operations. A team can use an assistant to inspect instance details, review networking configurations, or identify backup and volume information without automatically granting the ability to create, modify, stop, or delete resources.

GreenNode CLI for vServer automation

The grn command-line interface now supports the full vServer lifecycle. Users can create servers, configure networks, manage volumes, and work with security groups directly from the terminal instead of relying on the Portal for each action.

CLI access is especially useful for teams that need repeatable infrastructure workflows. It can be used in shell scripts, internal automation tools, infrastructure runbooks, and CI/CD pipelines where manual portal operations would be slow or difficult to standardize.

Together, GreenNode CLI and GreenNode MCP offer two complementary operating models:

WorkflowBest suited forPrimary interaction
GreenNode CLIRepeatable, scriptable, pipeline-based infrastructure tasksTerminal commands and automation scripts
GreenNode MCPContext-driven exploration and AI-assisted operationsNatural-language requests through compatible AI tools
GreenNode PortalVisual configuration and administrationWeb interface

The choice depends on the task. Use the CLI when predictable automation is required; use MCP when natural-language interaction and contextual assistance can accelerate investigation or workflow execution.

Backup, monitoring, and security updates

VBackup: Clearer schedules and missed-backup alerts

VBackup now displays the precise next backup time directly in the resource list. Users no longer need to calculate the next execution time manually from the assigned backup policy. When a policy changes, the Next backup time updates automatically for all servers and databases using that policy.

The platform also sends email alerts when a scheduled backup does not run at its expected time. This helps operations teams identify missed backup executions sooner, before a protection gap becomes a larger recovery risk.

VBackup additionally integrates with GreenNode MCP, providing more than 68 tools for managing backup policies, scheduling operations, and monitoring restore-related workflows through natural-language interactions.

For teams responsible for business continuity, these changes improve visibility across three critical questions:

  • When will the next backup run?
  • Did a scheduled backup fail to execute as expected?
  • How can backup and restore workflows be checked or managed without switching constantly between interfaces?

vMonitor Datasource Plugin for Grafana

GreenNode now provides a Grafana datasource plugin for vMonitor. The plugin connects Grafana directly with the vMonitor Platform, allowing users to query vMonitor metrics and logs, create dashboards, and configure alerting based on vMonitor data.

This is valuable for teams that already use Grafana as a centralized observability layer. Rather than separating infrastructure monitoring views across multiple tools, they can bring vMonitor telemetry into shared operational dashboards alongside data from other systems.

Typical use cases include:

  • Building dashboards for infrastructure health, application performance, and resource utilization.
  • Correlating vMonitor logs and metrics during incident investigation.
  • Creating Grafana alerts based on vMonitor data.
  • Giving engineering, SRE, and operations teams a common monitoring interface.

vWAF: DNS configuration through CNAME

vWAF now supports DNS configuration using CNAME records in addition to A records. GreenNode provides a dedicated DNS domain for each customer, enabling more flexible traffic routing between vWAF zones without requiring customers to update their DNS configuration whenever routing changes are needed.

This can simplify operational changes for teams that need to adjust routing while minimizing repeated DNS updates on the customer side. If a DNS provider does not support CNAME records for a root domain, users can continue using A records as before.

vDB: PostgreSQL Cluster now in Hanoi

PostgreSQL Cluster is now available in the Hanoi region. The service provides a multi-node architecture with automatic failover and separate read/write endpoints, bringing managed PostgreSQL cluster capabilities closer to workloads located in Northern Vietnam.

Separate read and write endpoints can help applications direct workload types more appropriately, while automatic failover improves resilience when a node becomes unavailable. For organizations running latency-sensitive applications, databases, or services in the Hanoi area, regional availability can also support architecture decisions that keep compute and database workloads geographically closer together.

What these updates mean

August’s updates make GreenNode more accessible through the interfaces technical teams already prefer:

  • AI assistants and MCP for conversational, context-aware operations.
  • CLI tools for repeatable automation and integration into engineering pipelines.
  • Grafana for centralized observability.
  • Improved platform visibility for backup scheduling, security controls, DNS routing, and database availability.

For platform engineering and DevOps teams, the practical opportunity is to reduce context switching. AI agents can connect to workplace systems, infrastructure can be operated from the terminal or through MCP-enabled assistants, and monitoring data can move into existing Grafana workflows.

Technical documentation

Explore the following documentation and repositories to get started with the August 2026 updates:

Get started

Customers can log in to the GreenNode Portal to explore the new capabilities, including AgentBase integrations, GreenNode MCP, VKS MCP Server, GreenNode CLI, VBackup enhancements, vMonitor’s Grafana datasource plugin, vWAF CNAME configuration, and PostgreSQL Cluster in the Hanoi region.

For architecture consultation or deployment support, contact the GreenNode team.

Frequently Asked Questions

What is GreenNode MCP?

GreenNode MCP is an integration based on the Model Context Protocol that lets compatible AI assistants and custom AI agents interact with GreenNode services through natural-language requests. For vServer, it includes more than 183 integrated tools covering instances, networking, volumes, snapshots, security groups, and SSH keys.

Is GreenNode MCP allowed to change infrastructure by default?

No. GreenNode MCP is read-only by default. Write access must be granted explicitly before an AI assistant or agent can perform actions that modify infrastructure resources.

Which tools can AgentBase agents connect to?

AgentBase MCP Connectors support Microsoft 365 services including Teams, Outlook, and SharePoint; GitHub; Atlassian tools such as Jira and Confluence; and Monday.com. Additional tool connections can be requested according to business needs.

How does the VKS MCP Server help Kubernetes teams?

The VKS MCP Server lets users connect VKS with compatible AI assistants such as Claude, Cursor, and VS Code-based tools. Teams can use natural-language requests to work with clusters, node groups, and Kubernetes resources instead of relying only on manual Portal operations.

What changed in VBackup?

VBackup now displays the exact next backup time, automatically updates that time when a policy changes, and sends email notifications when a scheduled backup does not execute on time. It also integrates with GreenNode MCP for backup policy, scheduling, and restore-related operations.

Can I use vMonitor data in Grafana?

Yes. The vMonitor Datasource Plugin connects Grafana directly to vMonitor, enabling metric and log queries, dashboard creation, and alerting workflows based on vMonitor data.

Does vWAF still support A records?

Yes. vWAF now supports both CNAME and A Record configurations. CNAME can provide greater routing flexibility, while A records remain available for environments where a DNS provider does not support CNAME records at the root domain.

Where is PostgreSQL Cluster now available?

PostgreSQL Cluster is now available in the Hanoi region, in addition to existing availability. It provides a multi-node architecture, automatic failover, and separate endpoints for read and write workloads.

TagProduct updateVKSCPU instancesGPU InstancesAgentBasepostgreSQL