Who is this on-call service for?
When you entrust your infrastructure to RDEM Systems, you want to know exactly what happens when an incident occurs at 3 a.m. This article explains in detail our organisation, tools, and on-call procedures. Full transparency.
1. Our on-call philosophy
Effective IT on-call support rests on three pillars:
Responsiveness
80% of alerts handled in under 45 minutes
Reliability
Automatic escalation -- no alert ever ignored
Documentation
Client runbooks accessible at all times
We adapt our coverage hours to your needs: full 24/7 for critical infrastructure, or after-hours only (AHO) if you already have a daytime team. Every contract is tailored to your operational reality.
2. PagerDuty: our nerve centre
We chose PagerDuty as our central incident management platform. It is the industry standard used by the largest tech companies (Netflix, Salesforce, IBM...) to orchestrate their on-call operations.
Why PagerDuty?
-
Centralisation
All alerts (monitoring + client emails) arrive in one place
-
Multi-channel
Simultaneous notification via phone call, SMS, and mobile push
-
Automatic escalation
If nobody responds, the alert automatically escalates to the next level
-
Full traceability
Every action timestamped: alert, acknowledgement, resolution
Triple-channel notification
When an alert is triggered, the on-call engineer receives simultaneously: a phone call, an SMS, and a push notification on the PagerDuty app. It is impossible to miss an alert.
3. How to trigger an alert
Two channels can trigger our on-call support:
Automated monitoring
Our monitoring automatically detects anomalies: server down, CPU saturation, critical disk space, stopped service, expiring certificate...
→ Instant trigger, no action required on your part
Client email
Noticed an issue not caught by monitoring? Send an email to your dedicated on-call address. The alert is created instantly in PagerDuty.
→ Dedicated email address provided at contract signing
In addition to our alerting matrix
PagerDuty notifications complement your own alerting matrix if you have one. We do not replace your internal alerts -- we supplement them with our on-call coverage.
4. 4-hour response time SLA escalation procedure
The 4-hour response time SLA is our standard offering, suited to important but non-critical infrastructure. Here is the exact process:
4-hour SLA timeline
Alert triggered
The on-call engineer receives phone call + SMS + push notification
1st re-notification
If no acknowledgement, the same engineer is notified again
Manager escalation
After 2 unanswered notifications, the team lead is alerted
Guaranteed response
An engineer is actively working on the incident
In practice
60% of alerts are handled within 30 minutes, 80% within 45 minutes. Manager escalation remains rare, but it guarantees that no alert slips through the cracks.
5. 1-hour response time SLA escalation procedure
The 1-hour response time SLA is our premium offering for critical infrastructure (e-commerce, SaaS, mission-critical business applications). The key difference: two engineers are on call simultaneously.
1-hour SLA timeline -- Dual on-call
Engineer #1 alerted
First on-call engineer notified immediately (phone call + SMS + push)
Engineer #2 alerted
If no acknowledgement from #1, the second engineer is notified in parallel
Guaranteed response
One of the two engineers is actively working on the incident
Why two engineers?
The dual on-call setup eliminates human risk: phone on silent, poor reception area, technical issue... With two people in parallel, the probability that neither responds is virtually zero.
| Criteria | 4-hour SLA | 1-hour SLA |
|---|---|---|
| On-call engineers | 1 | 2 |
| Re-notification | Every 30 min | Engineer #2 at M+30 |
| Manager escalation | After ~1h | After ~1h |
| Recommended for | Standard infrastructure | Critical infrastructure |
Not sure which level to choose?
We can help you size your on-call needs in a 15-minute call.
6. How an intervention unfolds
Once the alert has been acknowledged, here is how the intervention proceeds:
Acknowledgement
The engineer confirms they are handling the issue in PagerDuty. You are notified that someone is working on the problem.
Remote connection
Secure access to your infrastructure as per the agreed method: direct SSH, client VPN, or bastion host depending on your architecture.
Diagnosis
Incident investigation: logs, metrics, service status. Client runbooks consulted when available.
Resolution or escalation
Corrective action if within our scope (infrastructure, core services, application diagnosis). Otherwise, the issue is documented and escalated to your application teams.
Closure and report
The incident is marked as resolved in PagerDuty with a summary of actions taken. A detailed post-mortem is provided on request.
Scope of intervention
Our on-call service covers:
Included in on-call support
- • Infrastructure: servers, VMs, storage, networking
- • Core services: web, database, email, DNS
- • Security: firewall, certificates, access control
- • Application diagnosis (to the extent of our capabilities)
- • Guidance and recommendations if out of scope
Escalated to your teams
- • Business application bugs
- • Development / code fixes
- • Functional user issues
- • Application-specific configuration
Documentation and runbooks
Each client has a documentation space in our internal CRM. You can also grant us access to your own wiki so we can write specific runbooks. During on-call shifts, our engineers systematically consult this documentation before intervening.
7. Our shared on-call team
Our on-call support operates on a shared model: a pool of engineers covers all our clients. This model offers several advantages:
For you
- Optimised cost: you do not pay for a dedicated 24/7 engineer
- Diverse expertise: our engineers work across very different environments
- No human SPOF: if one engineer is unavailable, another takes over
For our engineers
- Distributed workload: rotations that allow genuine recovery time
- Less fatigue: fewer alerts per person = better responsiveness
- Skills growth: exposure to a wide range of technologies
Confidentiality guaranteed
Every engineer is bound by strict confidentiality clauses. Access to client environments is logged and auditable. We can provide access reports on request.
8. Our organisation: full transparency
A hands-on company director
The director of RDEM Systems holds the status of corporate officer (president of a French SAS company). This status means:
- No statutory working hours limitation
- No on-call constraints under labour law
- Full flexibility in organising work schedules
In practice: the director can intervene at any time without the regulatory constraints that apply to employees. This provides a reliable first line of response.
A networked team
RDEM Systems relies on a team of partner contractors to provide 24x7 coverage.
Why external contractors?
- No disguised employment: our contractors work for multiple clients, not exclusively for RDEM Systems
- Own responsibility: each contractor manages their own regulatory obligations for the shifts they cover
- Specialised expertise: each contributor brings specific skills (systems, networking, security...)
Ongoing internalisation: we are working to bring part of the team in-house to strengthen our response capacity and deepen our expertise on client infrastructures.
9. Legal compliance
Corporate officer status
Not subject to working-hours and on-call constraints that apply to employees
Independent contractors
Work for multiple clients, no exclusive subordination = no disguised employment
Delegated responsibility
Each contractor manages their own regulatory obligations for their intervention shifts
B2B commercial contract
You sign a service agreement, not an employment contract. Labour law obligations do not apply to you.
10. What this means for you
Zero regulatory burden
No advance notice periods, no compensations to calculate, no documents to produce. We handle everything.
1-hour response time SLA, contractually guaranteed
Our organisation allows us to guarantee intervention within 1 hour, 24 hours a day, 7 days a week, 365 days a year.
Reliability through redundancy
Dual on-call with automatic escalation. If the first person does not respond, the second takes over.
Multidisciplinary team
Access to systems, networking, security, and infrastructure expertise through our partner network.
11. FAQ
How much does downtime cost you?
Calculate the real financial impact of one hour of downtime
Case study: E-commerce SMB, 12 servers
Before
- On-call handled by 2 developers in rotation
- Average night-time response: 45 minutes
- 3 major incidents undetected in 6 months
- Burnout and turnover in the tech team
After (with RDEM)
- 24/7 on-call with guaranteed 4-hour response time SLA
- Average acknowledgement time: 8 minutes
- 0 undetected incidents since deployment
- Team refocused on product, not firefighting
Our packages with on-call included
24/7 on-call support via PagerDuty is included in every managed services package. Offsite backup via NimbusBackup PBS is also included.
Essential
70 EUR/mo
7am-10pm 7d/7 + 250 GB backup
Pro
150 EUR/mo
24/7 4-hour SLA + 500 GB backup
Critical
250 EUR/mo
24/7 1-hour SLA + 1 TB backup
Per server, excl. VAT. All hosting providers: OVHcloud, Hetzner, Scaleway, Contabo, IONOS.
View our 24/7 on-call plans (fr)Need on-call support for your infrastructure?
Get in touch to discuss your requirements. We will tailor an offering to your needs: 4-hour or 1-hour response time SLA, 24/7 or after-hours coverage, with or without monitoring included.
Related articles
How much does IT on-call support cost?
Pricing, billing models, and comparisons
SLA, SLO, SLI: the complete guide
Understanding and negotiating your service commitments
Outsourcing your 24/7 on-call support
Benefits, organisation, and selection criteria
On-call regulations in France (fr)
Labour code and legal obligations