# Faster Incident Recovery for a Digital Asset Exchange

> TechSupport Team took the Sev-2 and Sev-3 on-call rotation for VaultChain’s Go, Rust and Kafka trading platform. Over six months, the client’s PagerDuty records show mean time to restore for Sev-2 incidents falling from 38 to 11 minutes.

- Publisher: TechSupport Team
- Canonical page: https://techsupport-team.com/case-studies/digital-asset-exchange-incident-recovery/
- Updated: 2026-11-16

Client: VaultChain · Fintech & Payments · Singapore digital-asset exchange

## Client and engagement facts

- **Client:** VaultChain
- **Industry:** Fintech & Payments
- **Business context:** Singapore digital-asset exchange
- **Delivery model:** Managed on-call for Sev-2 and Sev-3
- **Application:** Trading platform with Go/Rust matching engine and Kafka pipeline
- **Supplied geography:** Singapore
- **Coverage:** 24/7, three shifts

## Application stack and tools

### Application stack

- Go
- Rust
- Kafka
- EKS

### Tools used

- PagerDuty
- Datadog

Related services: [24/7 technical support](https://techsupport-team.com/services/24-7-technical-support/)

## Why Support Was Needed

A weekend matching-engine stall paged seven people without a clear first action. The exchange’s infrastructure lead retained Sev-1 incident command. The managed rotation gave Sev-2 and Sev-3 response a named owner.

## Technical Work

- Built a runbook that started with 'check partition distribution first' after a 7-minute diagnosis.
- Added a pre-session check for consumer-group stability 15 minutes before each trading session.
- Documented every Sev-2 incident with timeline, evidence trail and prevention step.

## Representative Incident

A broker restart at 09:12 left Kafka partitions imbalanced. When the engine stalled at 09:14, the on-call engineer paused the consumer and rejoined it cleanly. Flow returned at 09:21 Singapore time. The team added a pre-session stability check.

## Results and how they were measured

Table: Reported results and their measurement scope

| Metric | Baseline | Result | Units | Measurement window | Sample / population | Source and scope |
| --- | --- | --- | --- | --- | --- | --- |
| MTTR (Sev-2) | 38 min | 11 min | min | 6 months |  | Client PagerDuty |
| Sev-1 incidents/quarter | 9 | 2 |  | 6 months | Platform Sev-1 frequency; outside the managed Sev-2 and Sev-3 response scope. | Client PagerDuty |
| Uptime | 99.92% | 99.99% | % | 6 months |  | Client Datadog |

## Evidence and Limitations

Two market-wide upstream outages were excluded. Sev-1 frequency is a platform outcome tracked by the client and sits outside the managed Sev-2 and Sev-3 response scope.

## Incident Command by Severity

VaultChain’s infrastructure lead retained Sev-1 incident command. TechSupport Team handled Sev-2 and Sev-3 response.

## Scope a Similar Engagement

List the incident severities, hours and trading windows that need a named on-call owner. [Request a support plan](https://techsupport-team.com/contact/) or explore [24/7 technical support](https://techsupport-team.com/services/24-7-technical-support/).

## Related pages

- [24/7 technical support](https://techsupport-team.com/services/24-7-technical-support/): The related service for this engagement.

## Contact

Request engineers from TechSupport Team or ask a question: https://techsupport-team.com/contact/
