Edit page

ESB — Enterprise Service Bus

Title

Slug

Status

Context

Categories

Topics

Markdown

# ESB — Enterprise Service Bus

## Status
- **Handover planned:** Before end of June 2026
- **Migration:** Dev team migrating to Azure API Manager — they will take over the application
- **Current owner:** DS team (Carlos)

## Tech Stack
- **Container:** JBoss Fuse
- **API management:** WSO API Manager (API endpoints)
- **Database:** MySQL — 3 nodes, clustered

## Architecture
- Java-based middleware
- Sits between systems as integration layer
- WSO handles external API endpoints

## Servers
- Located at: (TBD — on-prem RHEL, see ST items)
- Environment: PROD + DR (alignment gap known issue)

## Key Contacts
- Dev team: handling Azure API Manager migration
- DS team (Carlos): current operational owner

## Known Issues
- ESB DR alignment gap (PROD vs DR not in sync)
- Log4j vulnerability — closed with justification (Azure migration planned)
- SCOM alerts for REST API services unavailability (IM-1-12591273206)

## Open Tasks
- Alert mechanism in ESB (task-2746e915e7) — in sprint W13

## Knowledge Gaps
- [ ] List of PROD server hostnames
- [ ] ESB DR gap — what exactly is out of sync?
- [ ] Most common incidents and resolution steps
- [ ] WSO API Manager version

## Handover Status (updated 2026-03-26)
- Zero DS involvement planned post-handover, but not fully decided yet
- Azure API Manager migration ongoing — dev team leads

## Upstream connections
- Transparent to DS — not worth documenting fully given handover timeline

## SCOM Monitoring (as of 2026-03-27)

### Current active monitors
- Server health (general)
- CPU / disk
- TCP check on api.un.org:443

### Known gap
- ESB API Services panel in SCOM is **empty** — old API monitors were removed and never replaced
- Only app-level check is TCP port → does NOT confirm the application is actually working

### Proposed fix (pending dev team response)
- Option A: Dev team points to an existing meaningful API endpoint to monitor
- Option B: Dev team creates a lightweight  endpoint returning HTTP 200 OK
- Next step: Saurav (dev team) to confirm which endpoint to use
- Once confirmed: configure in SCOM, possibly open RFS

### Related SCOM dashboards
- Critical Applications dashboard
- ESB dedicated dashboard

## Azure API Manager (APIM) Rollout
- **Status:** In progress — dev team migrating APIs one by one
- **Target completion:** End of June 2026
- **Contact:** Saurav Datta (ESB developer)

## PMDS 2026 Objective (added 2026-04-01)

**Objective:**
Support the handover of the ESB platform to the development team and contribute to infrastructure cost optimization, ensuring a clear transition path by end of cycle.

**Planned Outputs:**
- Support infrastructure cost optimization by assisting with resource consolidation, removal of unnecessary servers, and migration to Proxmox to reduce operational costs.
- Collaborate with the development team to define and document the ESB handover process, covering infrastructure scope, responsibilities, and transition milestones.
- Contribute to the migration of remaining API endpoints and post-migration infrastructure definition in coordination with the development team.
- Ensure all prerequisites for removing ESB from the support catalog are identified, tracked, and progressed toward completion.

**Notes:**
- PMDS cycle ends May 13, 2026 (mid-term review); full handover target is end of June 2026
- API migration will not be complete by May 13 — outputs are framed as contributions/in-progress
- Carlos's role: supporting (not leading) both handover and cost rationalization

## Cost Rationalization & Proxmox Migration (added 2026-04-01)

- **Parallel effort** running alongside the Azure API Manager migration
- Goal: reduce operational costs — consolidating resources, removing unnecessary servers
- VMware → **Proxmox** migration underway (cost-driven, not part of Azure migration track)
- Some servers already decommissioned (Jan–Apr 2026)
- Important distinction:
  - **Partial decommission** = cost rationalization (remove unused infra now)
  - **Full decommission** = end of migration (all on-prem ESB infra gone post-migration)
- Carlos supports this effort; does not own cost targets or manage the rationalization project

Cancel