Upgrade Best Practices: DX Operational Observability (DXO2) version 24.1 to 26.1

Products

• DX Operational Observability

Issue/Introduction

You need to upgrade DX Operational Observability (DXO2) from version 24.1 to 26.1 and require best practices for sizing, infrastructure, and the sequential upgrade path.

Environment

• Product: DX Operational Observability (DXO2) / APM
• Platform: OpenShift Container Platform (OCP)
• Current Version: 24.1
• Target Version: 26.1

Resolution

You must follow a sequential upgrade path and ensure the underlying OpenShift cluster is updated to the supported version before each application upgrade step.

1. Sequential Upgrade Path

You must perform the upgrade in two main phases:

• Phase 1: Update your cluster to OCP 4.18, then upgrade DXO2 from 24.1 to 25.1.
• Phase 2: Update your cluster to OCP 4.20, then upgrade DXO2 from 25.1 to 26.1.

2. Required Backups

Before you start each upgrade phase, you must perform the following backups:

• Cluster Objects: Run the command ./AIOpsManagement.sh --backup to save deployments, configmaps, and helm charts. Note that this command does not save application data.
• Application Data:
  • Manually back up the entire NFS folder (includes Postgres, NAS, TAS, and ACC DB).
  • Create an Elasticsearch Cluster Snapshot.
  • Back up any manually created Javascript calculator scripts.

3. Infrastructure and Sizing Recommendations

For a medium-scale environment (supporting approximately 4,000 agents), use the following infrastructure configuration. You should prefer Bare Metal for worker nodes to ensure optimal performance.

 

 

Role	Node Count	CPU	RAM	Storage Details
Master	3	8C	16GB	VM
Infra Worker	3	8C	32GB	VM
Worker	7 + 2 (HA)	16C	64GB	Shared NFSv4.1 (10 TB) on Bare Metal

4. Storage Configuration Best Practices

• Use NFS Only: You should use NFS or Network storage for all components, including Elasticsearch and Kafka. This approach is recommended over LocalStorage for operational efficiency in OpenShift environments.
• Validate Performance: Before installation, you must validate your NFS IOPS and throughput. Use the dxi-fio-docker.sh or dxi-fio-test.sh scripts available in the installer package to test disk performance.
• Persistent Volumes: You should let the installer create Persistent Volumes (PVs) if your user has the storage admin role, or pre-create them manually using provided helper scripts. Note that auto-provisioning operators like OpenShift Data Foundation (Ceph) are not officially supported.

Additional Information

Techdocs DX O2 OnPrem installation