Meshly Data Stack / Deployment

We scope your footprint with you.

Meshly Data Stack is built for Red Hat OpenShift, and Meshly is a Red Hat partner. The stack runs on your infrastructure, and the Meshly team scopes and onboards it together with you.

The configuration tier decides which services run, and hardware sizing is a separate choice, because the same code runs at any size, from a laptop to a production rack. Meshly Data Stack is available to customers only, and we license and scope it in a consultation.

SYSTEM / AGENT OPERATIONS LAYER4 LAYERS

Four layers, from top to bottom. Your infrastructure is highlighted.

  • 4 · Agents and modelsCoding · Process · Local · CLIReplaceable workers, from any vendor.
  • 3 · Meshly BuildRules · Review · RecordNo agent approves its own work.
  • 2 · Meshly Data StackYour data · Meaning · PermissionAgents see only what they are allowed to.
  • 1 · Your infrastructure (highlighted)On-prem · Private cloud · Airgapped
Meshly Data Stack runs on your infrastructure: on-prem, private cloud or airgapped.FIG. 1
CustomStandardEnterprise32 to 37 containersOpenShiftKubernetesPodman

Detail 01, configuration tiers

You choose from three configuration tiers.

We recommend Standard for most teams. Enterprise adds stream and batch processing, and Custom lets you pick individual components when you need a smaller footprint.

TierWhat runsContainers
Tier: CustomBring your own footprintPick individual components · Keycloak and OPA always included · For tight resource budgets and specialised deployments · No stream or batch processing by defaultContainers: You pick
Tier: Standard (recommended)Full SSO and AILakehouse: Iceberg on MinIO with Nessie · Trino federation across all sources · Real-time CDC: Debezium to Kafka to any sink · Airflow, OpenMetadata and Superset · Keycloak SSO, OPA RBAC and OpenBao · Meshly Build connector and MDS-BI builder · Qdrant vectors and Valkey cache · MDS Dashboard includedContainers: 32
Tier: EnterpriseAll components, stream and batchEverything in Standard · Apache NiFi visual ingestion · Apache Flink stream processing · Apache Spark batch and ML, with the Spark History server · Complex event processingContainers: 37

Faust switches off automatically when Flink is on.

Detail 02, hardware sizing

Size the hardware for the work you will do.

Hardware sizing is a separate choice from the configuration tier, so pick the row that matches what you will actually do with the stack.

SizingWhat it meansRAMCPUsDisk
Boot floorWhat it means: The stack starts and demos open; no real queriesRAM: 12 GBCPUs: 6Disk: 40 GB
Light devWhat it means: One user, ad hoc Trino queries, light DAGsRAM: 16 GBCPUs: 8Disk: 100 GB SSD
RecommendedWhat it means: Several users, real Trino and OpenSearch load, ingestRAM: 32 GBCPUs: 16Disk: 250 GB SSD
ProductionWhat it means: Concurrent users, steady ingest, retained dataRAM: 64 GB+CPUs: 24+Disk: 500 GB+ NVMe

Detail 03, platforms

The stack is built for OpenShift.

Red Hat OpenShift is the production target. Meshly is a Red Hat partner, and the stack ships ready for your cluster with default-deny network policies for every service.

The same install also runs on Kubernetes, on a single server, and on macOS, Linux and WSL2 with Podman for development. You run the same stack at every stage, so moving from one to the next never turns into a migration project.

Cluster
The stack runs on Red Hat OpenShift and on Kubernetes 1.19+.The manifests ship default-deny NetworkPolicies and Secret-driven credentials
Server
Server mode generates the Nginx and SSL configuration and a subdomain for each service.It runs on your own domain
Workstation
The stack runs on macOS, Linux and WSL2.Workstations use Podman, for development

Detail 04, onboarding

The Meshly team onboards you.

  1. 01 / Step

    Talk to us

    Tell us about your current data setup, and we will assess it with you in a deployment conversation.

  2. 02 / Step

    Plan the configuration

    Pick the tier and the hardware sizing that fit your data and your team. We scope the licence at this step.

  3. 03 / Step

    Deploy and connect

    The Meshly team deploys the stack on your infrastructure and connects your first integrations.

  4. 04 / Step

    Run it from the portal

    Your licence, team, contracts, support and release updates all live in the customer portal.

Detail 05, customer portal

Your stack has its own portal.

The portal gives you self-serve access to your Meshly Data Stack, with your licence, team, contracts, contacts, support and release updates in one place. From licence to launch, you can see what is yours without filing a ticket.

Customers
You see your stack, your team and your contracts with one sign-in.The portal shows your team, your contracts and your contacts
Partners
Every customer you onboard gets their own tenant.Each customer keeps their own users and their own history, so you stay in control while they stay isolated.

Related sheets

Scope your deployment with us.

Tell us about your data, your team and your infrastructure, and we will help you pick the tier and the sizing and plan the onboarding.

One level deeper

Check the boundary yourself.

The Trust Center sets out the deployment boundaries, our security practices, and what is confirmed and what is still in progress.

Open the Trust Center