# FM-KED-000 Getting Started
---
The **Finmars Known Error Database** is a curated registry of recurring and well-understood operational issues observed in Finmars environments, together with documented recovery procedures and practical guidance.
Its purpose is not to list every possible failure, but to capture **patterns that are already known**, reproducible, and solvable through established algorithms. Each entry describes the symptoms, scope of impact, severity, and an expected recovery path based on real operational experience.
This document serves three audiences at once
- operators who need fast orientation during incidents
- customers who want transparency and predictability
- support teams who require a shared operational memory
Every known error in this database is classified by **severity** and **recovery class**, allowing readers to understand both the business impact and the typical effort required to restore normal operation. Where possible, step-by-step recovery instructions are provided to reduce response time and avoid unnecessary investigation.
Issues that are **not present** in this database are considered unknown by definition. Such cases may require discovery, diagnostics, or architectural analysis and are handled outside the scope of predefined recovery commitments.
The Finmars Known Error Database is a living document. It evolves with the platform, operational experience, and lessons learned from real incidents. Its goal is clarity, not completeness. Reliability, not illusion.
Each known issue carries **two orthogonal markers**:
1. **Severity** – impact on the business or system
2. **Recovery Class** – expected effort and time to restore service
---
### Severity Levels (Impact-based)
**S1 — Critical**
- System unusable or data integrity at risk
- No viable workaround
- Immediate attention required
**S2 — High**
- Core functionality degraded
- Workarounds exist but are painful
- Business impact is noticeable
**S3 — Medium**
- Partial feature loss
- Clear workaround available
- Limited operational impact
**S4 — Low**
- Cosmetic or non-blocking behavior
- No impact on business operations
Severity answers the question:
**“How bad is this for the user?”**
---
### Recovery Class A — Quick Fix
- Expected resolution: **under 1 hour**
- Fully documented procedure
- Typical for routine maintenance issues
### Recovery Class B — Standard Recovery
- Expected resolution: **up to 4 hours**
- Known issue with multiple steps or validation
- Usually fits within monthly support window
### Recovery Class C — Extended Recovery
- Expected resolution: **up to 1 business day**
- High complexity or cross-component dependency
- Not fully covered by standard support
### Recovery Class D — Exploratory
- Resolution time **unknown**
- Partial or missing diagnostic data
- Discovery and investigation required
Recovery class answers the question:
**“How long does it usually take when things go normally?”**
---
## How it looks in registry
**Example**
- **Issue ID:** FM-KED-###
- **Title:** Background jobs stuck in pending state
- **Severity:** S2 — High
- **Recovery Class:** B — Standard Recovery
- **Estimated Recovery Time:** up to 4 hours
- **Covered by Monthly Support:** Yes
- **Resolution Algorithm:** documented
For an unknown issue:
- **Severity:** S1 — Critical
- **Recovery Class:** D — Exploratory
- **Covered by Monthly Support:** No
- **Notes:** subject to separate agreement
---