Transition Management: Coexistence Strategies

Large-scale cloud migrations do not happen overnight. Government departments will operate in a hybrid state for 2-5 years, with systems running simultaneously on US cloud providers and sovereign infrastructure. This page addresses the practical realities of this coexistence period.

1. Projects In-Flight Decision Framework

At any given moment, government departments have dozens of cloud projects at various stages: procurement, development, testing, or recently launched. Each requires a decision about how to proceed.

Project Status Categories

Decision Matrix: What To Do With Each Project

2. Extended Parallel Operation Patterns

For complex systems, the transition period may extend 12-36 months. During this time, the system operates in both environments simultaneously. This is not a bug—it's a feature that reduces risk and enables gradual confidence building.

Parallel Operation Models

Extended Parallel Operation Timeline

MONTH    1    3    6    9    12   18   24   30   36
         │    │    │    │    │    │    │    │    │
US CLOUD ████████████████████████████████░░░░░░░░░░  (100% → 0%)
         │    │    │    │    │    │    │    │    │
SOVEREIGN ░░░░░░░░░░░░░░████████████████████████████  (0% → 100%)
         │    │    │    │    │    │    │    │    │
         │    │    │    │    │    │    │    │    │
PHASE:   │PREP│PILOT    │RAMP-UP    │PRIMARY      │COMPLETE
         │    │         │           │             │
DATA:    │    │ Read    │ Bi-dir    │ Sovereign   │ US
         │    │ Replica │ Sync      │ Primary     │ Decomm
        

3. Data Synchronisation Strategies

Maintaining data consistency across two cloud environments is the most technically challenging aspect of extended parallel operation. The strategy depends on data characteristics and consistency requirements.

Synchronisation Patterns

Change Data Capture (CDC) Implementation

CDC is the recommended pattern for most database synchronisation scenarios. It captures changes at the database level and streams them to the target environment.

# Example: PostgreSQL CDC with Debezium to sovereign infrastructure

# 1. Source Database (AWS RDS) - Enable logical replication
ALTER SYSTEM SET wal_level = logical;
ALTER SYSTEM SET max_replication_slots = 4;
ALTER SYSTEM SET max_wal_senders = 4;

# 2. Debezium Connector Configuration (runs in sovereign Kubernetes)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: cdc-source-connector
spec:
  class: io.debezium.connector.postgresql.PostgresConnector
  config:
    database.hostname: source-db.xxx.eu-west-2.rds.amazonaws.com
    database.port: 5432
    database.user: cdc_user
    database.password: ${CDC_PASSWORD}
    database.dbname: production
    database.server.name: aws-source
    plugin.name: pgoutput
    slot.name: debezium_slot
    publication.name: dbz_publication
    # Route through secure tunnel - NOT public internet
    database.sslmode: verify-full

# 3. Sink Connector (writes to sovereign PostgreSQL)
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: cdc-sink-connector
spec:
  class: io.confluent.connect.jdbc.JdbcSinkConnector
  config:
    connection.url: jdbc:postgresql://sovereign-db:5432/production
    connection.user: app_user
    topics.regex: aws-source.*
    insert.mode: upsert
    pk.mode: record_key
    auto.create: false
    auto.evolve: false

Conflict Resolution for Bidirectional Sync

When both environments can write, conflicts will occur. Define resolution rules upfront:

4. Contract & Commercial Management

Government departments have existing contractual commitments with AWS, Azure, and GCP—often multi-year enterprise agreements with committed spend. The transition must account for these commercial realities.

Contract Situation Assessment

Commercial Transition Strategies

Data Egress Cost Planning

Cloud providers charge for data leaving their networks. At government scale, egress costs can be substantial:

5. Industry Lessons Learned

Several major organisations have undertaken large-scale cloud migrations or repatriations. Their experiences provide valuable lessons.

6. Hybrid Steady-State: Systems That May Never Fully Migrate

Some systems may remain on US cloud infrastructure indefinitely due to technical, commercial, or practical constraints. This is acceptable if properly managed.

Candidates for Permanent Hybrid State

Managing Permanent Hybrid State

Hybrid Architecture Pattern

7. Operational Considerations During Transition

Monitoring & Observability

During coexistence, unified monitoring across both environments is essential:

• Single pane of glass: Grafana dashboards showing both environments
• Unified alerting: PagerDuty/Opsgenie routing regardless of source
• Distributed tracing: Jaeger/Tempo tracing requests across environments
• Log aggregation: All logs to sovereign infrastructure (even from US cloud)
• Synthetic monitoring: External probes testing both environments

Incident Response

Team Structure During Transition

Related Documentation

• Migration Patterns - Technical patterns for individual workload migration
• Emergency Migration Strategy - Accelerated timelines for crisis scenarios
• Risk Register - Risks associated with extended parallel operation
• Cost Model - Financial implications of dual-running
• Migration Strategy (Level 1) - Strategic overview