Cache hit
A query answered from cached data without contacting an upstream authoritative server. A high cache-hit rate can reduce latency and upstream load, though the right rate depends on workload and TTLs.
Category glossary
Search all termsOperations
26 definitions associated with Operations.
Cache miss
A query for which the resolver does not have usable cached data and must perform resolution or forwarding. Misses increase upstream traffic and usually add latency.
Capacity planning
Forecasting when address pools, prefixes, DHCP capacity, or DNS infrastructure will need expansion. Useful inputs include utilization, growth rate, lease churn, query rate, and failure-domain requirements.
Configuration drift
A difference between intended configuration and the state actually deployed across servers, appliances, cloud services, or data stores. Drift can create inconsistent answers, overlapping allocations, and failover risk.
CUDN
A centrally operated user-facing recursive DNS service, often used as an organizational resolver platform. The acronym is less universal than common terms such as recursive resolver.
DNS anomaly
A query or response pattern that differs from expected behavior, such as sudden NXDOMAIN growth, unusual labels, rare record types, high entropy, or traffic from an unexpected client. An anomaly is a signal, not proof of malicious activity.
DNS flag day
A coordinated industry event after which participating DNS software stops accommodating a legacy or non-compliant behavior. The goal is to accelerate protocol hygiene by changing defaults at the same time.
dnstap
A structured binary event stream for recording DNS message and resolver processing data. It provides richer and often more efficient telemetry than plain-text query logs.
EDNS buffer size
The UDP payload size a requester advertises in the OPT record. Oversized values can cause fragmentation or path failures, while undersized values cause more truncation and TCP or QUIC fallback.
Freeze
A product operation that temporarily stops dynamic modifications to a zone so its backing file or data can be edited safely. In BIND, thaw resumes normal dynamic update processing.
Garbage collection
Automatic cleanup of expired leases, stale records, orphaned objects, old logs, or unreferenced data. Cleanup policy should preserve audit history while preventing stale state from becoming operational truth.
ICMP echo
The request and reply messages commonly used by ping. Some DHCP servers use ICMP echo as one signal in IPv4 conflict detection, though lack of reply does not prove an address is unused.
Journal
A change log used by a DNS server to support dynamic update recovery and incremental zone transfers. Journal file formats and retention are implementation-specific.
Last known good
A previously validated configuration or data state retained for rollback or recovery. DDI changes should define how to restore a last-known-good zone, database, or service configuration.
NXDOMAIN spike
A sudden rise in nonexistent-name responses. Causes can include typos, missing suffixes, search-list behavior, stale applications, DGAs, random-subdomain attacks, or failed service discovery.
Ping
A diagnostic operation that sends ICMP echo requests and measures replies. Ping tests IP reachability, not DNS correctness or application availability, and devices may intentionally ignore it.
Probe
An active test used to measure reachability, latency, correctness, or service health. DNS probes may query specific names and validate answers, flags, timing, and server identity.
QPS
Queries per second, a common measurement of DNS request rate. Capacity planning should consider peaks, response sizes, cache behavior, query types, and failure scenarios rather than QPS alone.
Query logging
Recording DNS query metadata such as time, client, name, type, response code, latency, and selected policy. Logging improves troubleshooting and security visibility but requires privacy, retention, and volume controls.
RTT
Round-trip time, the elapsed time for a request to travel to a destination and a response to return. DNS RTT measurements should distinguish client-to-resolver and resolver-to-authoritative paths.
Service level agreement (SLA)
A formal commitment for measurable service outcomes such as availability, response time, support, or recovery. DDI SLAs should define measurement points and exclusions.
Service level objective (SLO)
A target for a service indicator, such as successful DNS responses, DHCP allocation time, or platform availability. SLOs are internal engineering objectives and may support an SLA.
Source of truth
The system or governed data set designated as authoritative for a class of information. In DDI, one system may own intended allocation while discovery or DHCP owns observed runtime state, so ownership must be explicit.
Synthetic monitoring
Automated tests that periodically perform DNS lookups, DHCP transactions, API calls, or application checks from controlled probes. It measures user-like outcomes even when no real user reports an issue.
Thaw
A BIND operation that resumes dynamic update after a zone has been frozen and reloads changes made to its master file.
Trap
An unsolicited management notification, commonly an SNMP trap, sent when an event or threshold occurs. DDI traps can report HA changes, pool exhaustion, service failure, or security events.