Monitoring and alerting
Health, capacity and availability monitoring on the agreed scope, with alerts routed to people who are actually on duty. We tune the thresholds so alerts mean something — an alarm nobody trusts is worse than no alarm.
CypherMax-delivered
Your infrastructure, looked after every day — not just when it breaks.
Monitoring, patching, backup verification, incident response and change management, run as a monthly service against a scope you have read and agreed to, and backed by a uptime commitment.
Fit
This is for you if
- Your alerts go to a channel nobody reads after 6pm.
- Backups are configured, but no one has restored from them.
- Patching happens when something forces it.
- Your one infrastructure-capable engineer cannot take a proper holiday.
- You need to answer a customer or investor security questionnaire and cannot evidence basic operational hygiene.
Scope
What is included
Patching and maintenance
Operating system and agreed platform-level patching on a defined cycle, applied to a staging environment first where one exists.
Backup and restore verification
We confirm backups are running, and we periodically restore from them. A backup that has never been restored is an assumption, not a backup.
Incident response
A defined escalation path and a named responder for incidents inside the agreed scope, with a written post-incident summary for anything significant.
Change management
Changes go through a documented request, review and rollback process. Small enough to not slow you down; formal enough that you know who changed what and when.
Access and identity hygiene
Periodic review of accounts, roles and keys — including the dormant admin account belonging to someone who left.
Monthly reporting
A monthly report covering availability, incidents, changes, patch status and spend, plus a quarterly review of what should change.
The commitment
What actually means
We commit to availability on environments under an agreed CloudOps scope. Because that number is worth very little without its definition, here is what sits behind it.
What is measured
The availability of the infrastructure and platform components inside the agreed scope, not your application code.
What it is not
A measured historical figure, or a commitment about a cloud partner's own platform availability. Partner platforms carry their own service commitments, which we will identify for your configuration.
The measurement window, the full exclusion list and the remedy are set out in the service agreement, and we will walk you through all three before you sign. An uptime figure without those three things attached is a slogan, not a commitment.
Boundaries
What is not included
Being clear about this up front is cheaper for both of us than discovering it during an incident.
- Application code. We operate the platform your application runs on. We do not debug your application logic, though we will help you find out which side a problem is on.
- Anything outside the agreed scope. The scope is written down and reviewed quarterly. Adding an environment is a conversation, not an assumption.
- A security programme. Operational hygiene is included; a formal security programme, penetration testing and certification audit work are separate.
- Unmanaged emergency work on environments we have not assessed. We cannot responsibly take on-call responsibility for an environment we have never seen.
Onboarding
How onboarding works
| Step | What happens |
|---|---|
| 1. Assessment | We review the environment and produce a findings document: what is healthy, what is at risk, and what must change before we can support it. |
| 2. Remediation | Agreed fixes to bring the environment to a supportable baseline — monitoring coverage, backup configuration, access control, documentation. Quoted separately, as a fixed scope. |
| 3. Scope agreement | A written service scope: environments covered, responsibilities on each side, escalation path, reporting cadence. |
| 4. Transition | Runbooks written, alerting routed, access provisioned, a handover session held with your team. |
| 5. Steady state | Monthly operation and reporting, with a quarterly review. |
FAQ
CloudOps FAQ
Usually not. More often we become their second pair of hands and their escalation path, so they can work on the things only they can do — and take annual leave.
Amazon Web Services, Google Cloud, Alibaba Cloud, Tencent Cloud, Huawei Cloud and BytePlus. If you run across more than one, that is normal and we will operate them together.
The least access that lets us do the agreed job, provisioned through named roles rather than shared credentials, and revocable by you at any time. We document exactly what we hold and why.
Yes, with notice.Yes. The notice period is set out in the service agreement. You keep your runbooks and documentation. An exit that leaves you unable to operate your own environment is not a service, it is a hostage situation.
Infrastructure and platform by default. Application-level monitoring can be added if you have the instrumentation, or built as a separate project if you do not.
Related