Back to all use cases

ConsultingIndustrial / Manufacturing

A dated plan for 19c: stay deliberately, or move with a rehearsed rollback

Industrial Manufacturing Group

A typical engagement to decide whether Oracle databases should stay on 19c or move to 26ai, and to plan the multitenant conversion, AutoUpgrade run and rollback for the ones that move.

Overview

This is a typical engagement for a manufacturer with a few dozen Oracle databases on 19c and 12c, some behind packaged applications, that wants a decision before Oracle's support dates force one. AceMQ and a partner Oracle DBA inventory the estate, test the real obstacles and write a plan the change board can approve.

Challenge

The right answer differs by database. Some are certified for 26ai by their application vendor and others are not. Some are non-CDB databases that must become pluggable databases as part of a 26ai upgrade. Some sit on platforms that Oracle's 19c support policy footnotes for Java 8. Oracle also supports direct upgrade to 26ai only from 19c and 21c, so the 12c databases need two steps.

Environment

A mix of Oracle Database 12.2, 18c and 19c on Linux and one other Unix platform, a number of non-CDB databases, packaged business applications and in-house systems, and a change calendar built around production shutdowns.

Approach

AceMQ classifies each database by release, edition, architecture, application certification and platform, then recommends stay, upgrade to 19c or upgrade to 26ai against Oracle's published dates. Each database that moves gets a rehearsal on a copy with the fallback tested.

Solution

  1. Inventories releases, editions, CDB or non-CDB status and options in use for every database
  2. Checks each application's vendor certification for 26ai before it enters the plan
  3. Maps Oracle's Lifetime Support Policy dates to each database and lists the Java 8 footnote cases
  4. Plans the AutoUpgrade runs, the non-CDB to pluggable database conversion and the order of the estate
  5. Rehearses the upgrade on a copy of production and records the downtime for each step
  6. Writes the rollback plan and keeps the platform change and the release change in separate windows

Outcome

The engagement delivers a per-database decision table, a dated upgrade plan with owners, rehearsal results for the first wave, a rollback plan for each step and a written statement of what staying on 19c requires in the meantime. The change board gets a plan it can approve instead of a single weekend with no way back.

Technologies

Oracle DatabaseAutoUpgradeMultitenantLinux
  • Assessment

    Oracle Data Guard Failover Readiness Review

    Regional InsurerCross-Industry

    A typical engagement to find out whether an Oracle Data Guard standby will really take over: transport and apply lag, protection mode, redo sizing, failover runbooks and client reconnection.

    Oracle DatabaseData GuardRMANLinux
    Read case study: Oracle Data Guard Failover Readiness Review
  • Managed Services

    Remote Oracle DBA Cover for a Small Platform Team

    SaaS Platform TeamSoftware

    A typical engagement where a lean platform team adds remote Oracle DBA cover: on-call, restore testing, patch windows and a monthly review, without hiring a second DBA.

    Oracle DatabaseRMANPrometheusLinux
    Read case study: Remote Oracle DBA Cover for a Small Platform Team
  • CVE Patching

    Closing the CVE-2023-46604 OpenWire Exposure Across a Mixed ActiveMQ Estate

    Industrial Equipment ManufacturerIndustrial

    A manufacturer with ActiveMQ Classic brokers at plants and in the cloud had no inventory of versions or clients when CVE-2023-46604 landed. AceMQ contained it, found what was exposed, and patched brokers and clients.

    ActiveMQJavaLinuxDocker
    Read case study: Closing the CVE-2023-46604 OpenWire Exposure Across a Mixed ActiveMQ Estate
  • Consulting

    ClickHouse Schema and Sort Key Design Consulting

    Industrial IoT Platform OperatorIndustrial

    Redesigning ORDER BY keys, partitioning, codecs, and materialized views so dashboard queries read a small fraction of the data instead of full scans.

    ClickHouseKafkaGrafanaLinux
    Read case study: ClickHouse Schema and Sort Key Design Consulting
  • Support

    Greenplum ETL Window Concurrency and Spill Support

    Global Consumer Goods ManufacturerIndustrial

    Diagnosing queries that queued and never started during the nightly ETL window, and spill files that filled segment disks, then separating workloads with resource queues and memory limits.

    GreenplumVMware vSphereLinux
    Read case study: Greenplum ETL Window Concurrency and Spill Support

Related Reading

Need Oracle Database Architecture Guidance?

AceMQ's senior Oracle Database engineers have handled this exact type of engagement before. Whether you need architectural guidance, hands-on remediation, or an ongoing managed partnership, we're ready to help.