English for Infrastructure Engineers: Vocabulary for Storage, Networking, and Ops

Master the English vocabulary infrastructure engineers use for storage systems, networking, incident response, and operational documentation.

Infrastructure engineers manage the hardware, networks, storage systems, and operational processes that keep services running. The vocabulary spans multiple domains — from SAN configurations to BGP routing to incident runbooks. This guide covers the core English terms you’ll encounter and use in infrastructure roles.

Storage Vocabulary

TermDefinition
SANStorage Area Network — a dedicated high-speed network that provides block-level storage to servers
NASNetwork Attached Storage — a file-level storage device accessible over a network
NVMeNon-Volatile Memory Express — a high-performance storage protocol designed for SSDs
TieringAutomatically moving data between storage tiers (e.g. fast SSD, slower HDD, archive) based on access frequency
IOPSInput/Output Operations Per Second — a measure of storage performance
ThroughputThe volume of data transferred per unit of time (e.g. GB/s)
SnapshotA point-in-time copy of a storage volume, used for backups and recovery
ReplicationCopying data to a secondary location in real time for redundancy

Storage in Practice

When discussing storage requirements, engineers typically specify both IOPS (for latency-sensitive workloads) and throughput (for large sequential reads/writes).

“The database requires a minimum of 10,000 IOPS with sustained throughput of 500 MB/s — the current SAN tier cannot meet this requirement under peak load.”

Networking Terms

TermDefinition
BGPBorder Gateway Protocol — the routing protocol that manages how packets are routed across the internet
VLANVirtual Local Area Network — a logical segmentation of a network
Load balancerA device or service that distributes incoming traffic across multiple servers
FailoverAutomatically switching to a secondary system when the primary fails
CDNContent Delivery Network — a globally distributed network of servers that caches content close to users
PeeringA direct network interconnection between two organisations
MTUMaximum Transmission Unit — the largest packet size that can be transmitted on a network
LatencyThe time for a packet to travel from source to destination

BGP in Plain English

BGP can be intimidating vocabulary. A simple way to explain it:

“BGP is the protocol that tells routers across the internet how to reach different networks. When we announce our IP range via BGP, we’re telling the internet ‘traffic destined for these addresses should come to us’.”

Incident Vocabulary

TermDefinition
MTTRMean Time to Recovery — average time to restore a service after an outage
RunbookA documented set of procedures for responding to a known operational issue
On-call rotationA schedule where engineers take turns being the primary responder to alerts
Escalation pathThe sequence of people to contact if the on-call engineer cannot resolve the issue
SLAService Level Agreement — a commitment to a minimum level of service
SLOService Level Objective — an internal target for service performance
Error budgetThe allowable amount of downtime defined by an SLO
PostmortemA structured analysis of an incident after it is resolved

Runbook Writing Language

Runbooks use imperative verbs and clear, numbered steps:

  1. “Verify that the service is unreachable by running curl -I https://api.example.com.”
  2. “Check the load balancer health in the AWS console and confirm that at least two instances are healthy.”
  3. “If fewer than two instances are healthy, escalate to the infrastructure lead.”
  4. “If the issue is a full outage, declare an incident using the incident management tool and notify the engineering manager.”

Discussing Infrastructure in Team Meetings

Capacity planning:

  • “Based on current growth trends, we’ll exceed our storage capacity in approximately five months — we should plan the expansion now.”

Incident reviews:

  • “The failover took 8 minutes, which exceeded our SLO. The runbook didn’t cover this failure mode.”

Architecture discussions:

  • “We should consider adding a CDN layer in front of the static assets — it would reduce origin server load by an estimated 60%.”

Example Sentences

  1. “The NVMe storage tier is reserved for the database primary — all read replicas use the standard SSD tier, which provides sufficient IOPS for our query patterns.”
  2. “BGP peering with our upstream provider was established successfully; we’re now announcing our /24 prefix from both data centre locations.”
  3. “Following the storage failover last Tuesday, we identified that the VLAN configuration on the secondary site was missing the correct routing rules.”
  4. “Our current MTTR for storage-related incidents is 45 minutes — the majority of this time is spent identifying the affected volume from the alert.”
  5. “The runbook for this failure mode was out of date; we’ve updated it and added an automated pre-flight check to the deployment pipeline.”