Calculate SLA in MyOperator (component-level method)

Applies to: Admins/Owners, Finance/Ops, and Support reviewers.
Outcome: Compute monthly uptime % for a specific component (e.g., Incoming — Delhi (TATA), Outgoing — Mumbai, IVR, SMS).
Time to complete: 5–10 minutes.

What “component-level SLA” means

MyOperator measures uptime per component (direction + service + region/provider). An incident on one component does not penalize others.

Prerequisites

• The component name you’re calculating for (e.g., Incoming — Delhi (TATA)).
• The billing month in IST (Asia/Kolkata).
• Access to Status Page/Incident log (or Dashboard export) showing: start time, end time, component(s), and labels such as No SLA Impact or Planned Maintenance.
• (Optional) Spreadsheet tool (Excel/Google Sheets) for formulas.

Step-by-step calculation

1. Define total minutes in month (IST)TotalMinutes=DaysInMonth×24×60\text{TotalMinutes} = \text{DaysInMonth} \times 24 \times 60TotalMinutes=DaysInMonth×24×60Sheets formula (copy-paste):
• Place any date of the month in A1.
• = (EOMONTH(A1,0) - EOMONTH(A1,-1)) * 1440
2. Collect incident windows for the chosen component
   From Status/Incidents, list all entries where this component is marked Degraded or Outage within the month.
• Exclude incidents tagged No SLA Impact (see section below).
• Exclude incidents tagged Planned Maintenance (inside the announced window).
3. Trim to the month boundaries (IST)
   For incidents spanning month edges, count only the overlap minutes within the month.
4. Sum downtime minutesDowntimeMinutes=∑(IncidentEnd−IncidentStart) in minutes\text{DowntimeMinutes} = \sum (\text{IncidentEnd} - \text{IncidentStart}) \text{ in minutes}DowntimeMinutes=∑(IncidentEnd−IncidentStart) in minutesSheets helper (per row):
• Start in B (Start) and C (End) as date-time (IST).
• Duration minutes in D: = (C2 - B2) * 1440
• Total: = SUM(D:D)
5. Compute uptime % for the componentUptime%=(1−DowntimeMinutesTotalMinutes)×100\text{Uptime\%} = \left(1 - \frac{\text{DowntimeMinutes}}{\text{TotalMinutes}}\right) \times 100Uptime%=(1−TotalMinutesDowntimeMinutes​)×100Sheets formula:
   = ROUND( (1 - (Downtime/Total)) * 100, 4 )
6. State the result clearly
   “Incoming — Delhi (TATA) uptime for July 2025 (IST): 99.9820% (Downtime: 7.8 minutes / Total: 44,640 minutes).”

Partial/low-impact handling

• Incidents labeled “No SLA Impact” are excluded from downtime tallies.
• These typically reflect minor or localized issues where measurable service levels remained within policy thresholds.
• You will still see them on the Status Page for transparency, but they do not reduce uptime%.

Edge cases & exclusions

• Planned maintenance within an announced window → excluded from downtime.
• Multi-component incidents: Count downtime separately per component affected. Do not add components together.
• Cross-month incidents: Count only minutes that fall inside the target month (IST).
• Intermittent flaps: Merge contiguous/overlapping windows before summing, to avoid double-counting.
• Data sources disagree: Prefer the official Status/Incident export; attach notes if you adjust times.

Troubleshooting & escalation

If your internal calculation doesn’t match ours, send:

• Component name + Month (IST)
• Your downtime tally (minutes) and the incidents you counted
• Export/CSV or screenshots showing start/end times and labels

Escalation template (copy-paste):
Subject: SLA calc check — [Component] — [Month, YYYY]
Body:

• Account/Company: [Name]
• Component: [e.g., Incoming — Delhi (TATA)]
• Month (IST): [e.g., July 2025]
• My tally: [Downtime minutes / Uptime %]
• Incidents included: [#IDs + times]
• Excluded as No SLA Impact / Planned: [#IDs]
• Notes: [overlaps/edge handling]

FAQ

Does an “Outgoing — North Region” issue affect “Incoming — North Region” SLA?
No. SLAs are computed independently per component.

How precise are times?
All calculations use IST timestamps; round minutes according to the incident export (no per-second rounding unless specified).

Can I publish an overall account SLA?
You may present a reporting roll-up (e.g., weighted by traffic), but contractual SLA remains per component.