System administration
Users, packages, disks, networking and the baseline the app depends on.
Linux Server Management & Support
Patching, observation and a named path when something breaks — on hosts you already pay for.
The management job
Own the host after go-live — with contracted hours, not a blank cheque.
Operations for a listed set of machines. Not unlimited on-call unless the agreement says so. Provider incidents stay with the provider.
- Agreed hostsThe list in the support agreement
- Patch and operateUpdates, services, disks
- ObserveLogs, capacity, known alerts
- RespondIncidents inside contracted hours
What is typically in scope
Administration work, not a vague “we handle IT”
Patching
Planned updates with service-impact awareness. Emergencies if the agreement covers them.
Service management
Web, database, queue and containers staying up and restarting cleanly.
Log analysis
Use logs to explain errors instead of restarting blindly.
Performance
CPU, memory, IO and slow queries when the app feels “just slow”.
Storage
Growth, inodes and backup footprint before the disk fills.
Database support
Access, backups, health and recovery for engines you already run.
Web server support
Nginx or Apache vhosts, TLS and common 5xx investigation.
Incident troubleshooting
Structured diagnosis when a site or API is down, in contracted hours.
Monitoring
Alerts mapped to hosts you care about.
How support is bounded
- A written host list — not “everything in the account”
- Coverage hours and a way to raise an incident
- Access limited to what the work requires, under your provider login
What we will not imply
- 24/7 cover unless that is contracted.
- That a provider outage becomes ADGLOB’s platform incident.
- That management is a cheap substitute for buying compute from us. You still pay the provider.
Related pages
List the hosts that need a named engineer
Bring the provider, the applications, and the hours you actually need someone to answer.
FAQ
Common questions
What does server management include?
System administration, patching, service management, log and performance review, database and web-server support, and incident troubleshooting within the agreed host list and hours.
Is this 24/7 incident cover?
Coverage hours and response targets are defined in the support agreement. We do not imply unlimited on-call unless that is contracted.
Do you need root access?
Usually yes, or an equivalent privileged role, for patching and incident work. Access is limited to what the work requires and remains under your account.
What if the cloud or VPS provider has an outage?
Platform incidents stay with the provider. We help you see the impact, recover the workload when the platform returns, and review whether the application can tolerate that class of failure.