How should a new computer science graduate start a career in systems administration?

Start with the fundamentals: TCP/IP, DNS, DHCP, VLANs, routing, and subnetting. If DNS or networking fails, nearly every service can appear broken, so understanding how packets move is essential.

Given your Linux and programming background, a useful progression would be Linux administration, containers, Ansible, and scripting or automation. Later, you can add AWS or Azure, Terraform, and eventually Kubernetes. Build a lab that grows in stages: begin with Linux virtual machines, deploy an application in containers, manage the configuration with Ansible, provision infrastructure with Terraform, and then explore cloud or container orchestration.

A CCNA is also worth pursuing if you have access to affordable training. Linux administration combined with automation can eventually lead naturally toward cloud engineering or DevOps roles.

A common entry point is Tier 1 help desk, but try to find an organization where you can actually learn and move upward. Some large support centers mainly involve password resets, while smaller teams or internal IT departments may let you work with servers, networks, virtualization, and automation. Keep building skills so you do not get stuck at the first level.

For certifications, CompTIA A+ and Network+ can provide a foundation. A CCNA or an entry-level cloud certification such as AZ-900 can help you stand out, although networking fundamentals matter more than collecting certificates. A home lab would also be valuable. Since you already have old servers available, experiment with Proxmox or Hyper-V and document what you build, what breaks, and how you fix it.

Do not overlook virtualization, storage, and the broader infrastructure around your servers. Proxmox and Hyper-V are useful technologies to practice, and containers are worth understanding even though they are not the answer for every workload.

For storage, learn the differences between local disks, DAS, NAS, SAN, and object storage. Study RAID, SMB, NFS, iSCSI, IOPS, throughput, latency, caching, and capacity planning. Also understand how storage connects to virtualization, backups, snapshots, replication, and disaster recovery. A snapshot is not a backup, and a system can have plenty of free space while still suffering from serious performance problems.

More generally, practice thinking about the consequences of every change: dependencies, security, access control, backups, recovery procedures, availability, and the effect on users. Reliable administration is much more than getting a server running once; it is making the result repeatable, maintainable, secure, and recoverable.

Your class was a good starting point, but systems administration is much broader than deploying a few communicating virtual machines. In real environments, you also need to consider reproducible builds, patching, security controls, permissions, monitoring, backups, high availability, rollback plans, business continuity, and disaster recovery.

A good lab project would deliberately include those concerns. For example, deploy a small service, automate its configuration, restrict access, monitor it, back it up, restore it, and document how you would recover after a failed update. That will teach you more than simply installing another server.

Start applying for roles now rather than waiting until you feel fully prepared. Interviews can show you which technologies employers actually expect; whenever a question exposes a gap, use it as a target for your next study project.

Managed service providers can be a useful, if demanding, first step because you may see many different environments and business setups. They can build a broad foundation quickly, but watch your workload and treat the role as a stepping stone. After gaining solid experience, look for an internal IT or dedicated systems position that better matches the direction you want.