KubeBlocks for
Milvus
Milvus is a cloud-native vector database designed for billion-scale similarity search. With GPU-accelerated indexing, hybrid search capabilities, and multi-tenancy support, it powers enterprise AI applications from semantic search to generative AI.
Supported versions
Available on
AWS
Azure
GCP
OCI
Alibaba Cloud
Rancher
OpenShiftDatabases
MySQL
PostgreSQLOracle
SQL Server
Redis
MongoDB
ClickHouseVector & AI
Qdrant
Milvus
ElasticsearchMessage queues
RocketMQ
RabbitMQ
KafkaOthers
VictoriaMetrics
InfluxDB
etcdExtend database engines like plug-ins
KubeBlocks provides unified database operations through its addon-based architecture. With KubeBlocks Enterprise, access over 15 seamless integrations to scale your database services.
One control plane, consistent operations across all engines — powered by the addon mechanism.
Operate Milvus with real distributed lifecycle, visibility, and recovery workflows
Run Milvus on Kubernetes with real console-backed evidence across lifecycle, scaling, observability, audit, and backup workflows, while keeping current console limits around access rules, parameters, collection management, and credential creation explicit.
Lifecycle
Provision Milvus and keep distributed service actions visible
Launch a tested distributed Milvus deployment, keep dependency-backed topology visible on the overview, and follow Restart Cluster progress from the same workspace.
- Create Milvus with the tested distributed topology and linked dependencies already wired into the cluster.
- Run Restart Cluster without leaving the overview and keep proxy restart progress visible in context.
- The same workspace also exposed Stop Cluster and Start Cluster for the tested deployment.
The overview keeps Milvus topology, dependencies, and lifecycle state visible right after provisioning.
Restart progress stays attached to the Milvus cluster so teams can watch distributed components recover in context.
Scale
Increase Milvus capacity and keep current topology limits explicit
Use Vertical Scaling to grow Milvus capacity from the overview and rely on the console to make missing rebalance or upgrade controls visible instead of implicit.
- Use Vertical Scaling to raise the tested proxy instance class from `1C1G` to `1C2G`.
- Track scaling progress from the same overview used for service state and distributed topology.
- No separate rebalance or upgrade workflow surfaced for the tested Milvus deployment.
Vertical Scaling keeps Milvus capacity growth inside a guided day-2 operation.
Parameter
Confirm whether in-console Milvus tuning is available
The Parameters route is reachable from the tested cluster and clearly shows whether runtime configuration is exposed for Milvus in the current console build.
- Review parameter coverage before planning Milvus tuning or template-driven configuration changes.
- Keep configuration review separate from lifecycle, scaling, and backup workflows.
- The tested Parameters page explicitly reported that this feature is unsupported for Milvus.
The Parameters route makes Milvus tuning coverage explicit instead of leaving teams to infer it.
Accessibility
Check connection-governance coverage from Access Rules
Access Rules keeps network-exposure review easy to find from the Milvus cluster and makes unsupported controls explicit when SSL and whitelist management are not available.
- Open SSL and IP Whitelist tabs from the same Milvus workspace used for lifecycle and scaling.
- Use the page as the operational source of truth for current connection-policy exposure.
- The tested console explicitly reported that this feature is unsupported for Milvus.
Access Rules keeps Milvus exposure review discoverable while making current feature limits clear.
Data Management
See exactly where collection and credential workflows stop today
The tested console surfaced the relevant Milvus routes, but collection setup, vector workbench handoff, and usable account creation were not fully exposed in-console.
- The Databases route explicitly reported that collection and index management are unsupported for the tested engine.
- The Credentials route loaded, but it did not expose an actionable account-creation flow in the tested session.
- No usable in-console vector write or search workbench surfaced from the Data Management route.
The Databases route shows current Milvus collection-management coverage clearly before teams plan application onboarding.
The Credentials route makes the current Milvus account-management boundary visible instead of hiding it behind a generic placeholder.
Observability
Track distributed Milvus health with metrics and runtime logs
Use Cluster Monitor to watch Milvus health across components and open Runtime Log to inspect real service output when troubleshooting or validating restarts.
- Monitor Milvus charts after Cluster Monitor finishes hydrating.
- Inspect real proxy startup and session-registration output from Runtime Log.
- Keep runtime visibility separate from task history and backup workflows.
Cluster Monitor gives teams a live view into Milvus health across distributed components.
Runtime Log captures real Milvus service activity so teams can validate restart behavior and investigate issues quickly.
Audit
Keep Milvus operational history traceable from Tasks
Task history records lifecycle and scaling changes so teams can confirm what happened before handing the cluster back to application users.
- Review Start, Stop, Restart, and Vertical Scaling from one operational timeline.
- Use task status as evidence that a Milvus day-2 action completed successfully.
- Keep change tracking separate from monitoring and data-protection workflows.
Task history gives teams a practical audit trail for Milvus lifecycle and capacity changes.
Backup
Protect vector data with real backup and restore workflows
Use the Backups page to capture protection points and open Restore from a completed backup whenever teams need validation or recovery options.
- Review completed full backups with repository, size, duration, and status in context.
- Launch restore directly from a real Milvus backup set instead of reconstructing recovery inputs manually.
- Keep recovery planning separate from routine lifecycle and scaling operations.
Backup history gives Milvus teams a real protection point they can rely on before upgrades, tests, or recovery drills.
The restore wizard turns an existing Milvus backup into a guided recovery path for validation and recovery planning.
Ready to build your own DBaaS on Kubernetes?
Talk to our team and see how KubeBlocks Enterprise can help you consolidate databases, strengthen security, and reduce operational costs.