NSEA flashcards
NSEA · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Hyperconverged infrastructure
- Architecture that combines compute, storage, and virtualization on the same x86 nodes, with a distributed storage fabric running as a controller VM on every host.
- Controller VM (CVM)
- Virtual machine on each Nutanix node that owns the local disks, runs the distributed storage fabric, and serves storage I/O to guest VMs.
- AHV
- Nutanix's native KVM-based hypervisor that runs guest VMs on each node and integrates with the distributed storage fabric.
- Prism Element vs Central
- Prism Element manages a single cluster, while Prism Central aggregates multiple clusters and provides categories, policies, and cross-estate reporting.
- Extent and extent group
- An extent is a 1 MB unit of data in the distributed storage fabric; an extent group is a 4 MB container holding multiple extents.
- Minimum cluster size
- A production Nutanix cluster needs at least three nodes for RF2 redundancy; RF3 designs require at least five nodes.
- Cassandra metadata store
- Distributed, replicated key-value database that tracks where every extent lives across the cluster and survives individual node failures.
- Stargate
- I/O mediation service running in the CVM that handles read and write requests from guest VMs and routes them to the correct extent location.
- Curator
- Background MapReduce-style service that performs disk balancing, scan, and re-replication of extents after a failure.
- Genesis
- Cluster formation and configuration service that elects the cluster leader and manages node join and removal operations.
- Prism Central
- Separate VM or VM set that aggregates data from multiple Prism Element clusters and enables categories, policies, and multi-cluster management.
- Quorum
- Requirement that a majority of metadata nodes be available; losing too many CVMs at once can make the cluster read-only or unavailable.
- Replication Factor
- Number of copies of each extent: RF2 tolerates one node or disk failure, while RF3 tolerates two at roughly 50% usable capacity.
- Effective capacity formula
- Usable capacity is approximately raw usable capacity divided by RF, minus reserved space and CVM overhead; never quote raw capacity as usable.
- Erasure coding
- Capacity-efficient data protection scheme that reduces overhead versus RF for cold data but is unsuitable for hot or latency-sensitive workloads.
- Protection Domain
- Grouping of VMs used for consistent snapshot and replication; a consistency group within it ensures multi-VM crash consistency.
- Async replication RPO
- RPO equals the replication interval plus lag; data written after the last successful replication can be lost in a disaster.
- Metro Availability
- Synchronous replication between sites requiring low-latency links, typically ≤5 ms RTT, providing near-zero RPO and automatic failover.
- Open vSwitch in AHV
- Virtual switch used by AHV to connect guest VMs to VLAN-backed networks and to the host bridge carrying management and storage traffic.
- Backplane network
- Network carrying CVM-to-CVM storage traffic; 10 GbE or higher per node is standard, with 25 GbE common for storage-heavy designs.
- Nutanix Flow
- Distributed firewall and microsegmentation capability applied at the VM vNIC level, so security policies follow the VM rather than the port.
- LACP bonding
- 802.3ad link aggregation that requires matching switch configuration and aggregates throughput; active-backup instead gives redundancy without aggregation.
- Jumbo frames
- MTU 9000 setting that must be configured end-to-end on the storage network; a mismatch causes fragmentation and performance degradation.
- Foundation
- Deployment tool that images nodes and forms a cluster from a configuration file describing node IPs, CVM IPs, and the cluster VIP.
- Cluster VIP
- Virtual IP address that fronts Prism Element for a cluster; the Prism Central VIP is separate and must be planned in DNS.
- Node expansion
- Process of racking and cabling a node, running the expansion workflow from Prism, and confirming the new CVM joins quorum before rebalancing.
- NCC
- Nutanix Cluster Check health-check framework; schedule regular runs and review failures before they escalate into outages.
- Lifecycle Manager
- Prism capability that manages AOS, hypervisor, firmware, and NCC updates; inventory compliance before scheduling any upgrade.
- Categories
- Key-value tags applied to VMs so protection, security, and placement policies can be applied by category rather than per VM.
- Replication lag
- Delay between data written on the primary cluster and its arrival at the DR target; it must be monitored to validate the achieved RPO.
- vCPU-per-core ratio
- Sizing assumption for how many virtual CPUs each physical core supports; required cores equal total vCPU divided by this ratio plus headroom.
- RPO and RTO
- Recovery Point Objective is the acceptable data loss window; Recovery Time Objective is the acceptable service restoration time after a failure.