NPS flashcards
NPS · 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.
- Controller VM (CVM)
- A service VM running on every Nutanix node that executes the AOS software stack and serves all storage I/O for its own host's VMs and for peer nodes.
- Genesis
- The AOS process manager on each CVM; it starts, monitors, and restarts the other Nutanix services such as Stargate, Cassandra, Prism, and Zookeeper.
- Stargate
- The AOS service that handles all vDisk I/O: it accepts reads and writes from the hypervisor, manages the extent store, and enforces data locality and checksums.
- Extent vs extent group
- An extent is the 1 MB unit of data that carries a checksum; an extent group is 4 MB made of four extents and is the unit used for placement and compression.
- Cassandra in Nutanix
- A distributed NoSQL database storing cluster metadata such as vDisk-to-extent mappings; every CVM runs a Cassandra node and the ring tolerates the loss of a node.
- Zookeeper role
- Stores cluster configuration and membership and elects leaders among AOS services; a quorum of Zookeeper instances must be available for cluster operations.
- RF2 capacity overhead
- With replication factor 2 every extent has two copies, so usable capacity is about 50 percent of raw, and the cluster survives one disk or node failure.
- RF3 fault tolerance
- Three copies of each extent let the cluster tolerate two simultaneous failures, while usable capacity drops to roughly 33 percent of raw capacity.
- EC-X erasure coding
- Erasure coding stripes data with parity across nodes to reduce overhead, for example 4+1 is about 1.25x; it applies only on RF2 clusters to cold data.
- Nutanix snapshot mechanics
- Snapshots are redirect-on-write metadata captures taken almost instantly; storage is consumed only as the source vDisk blocks are changed after the snapshot.
- NearSync RPO
- Nutanix NearSync replication lowers the recovery point objective to 1-15 minutes, whereas standard snapshot-based asynchronous replication has a minimum RPO of one hour.
- Curator
- The AOS service that runs background jobs including disk balancing, ILM data tiering, snapshot deletion, and reservation management to keep capacity and placement healthy.
- Cerebro
- The AOS service responsible for snapshots and replication to remote clusters; it coordinates protection domain policies and moves DR copies between sites.
- Open vSwitch in AHV
- AHV uses Open vSwitch as its virtual switch; the br0 bridge carries management traffic for the CVM and host, and guest networks attach as separate bridges.
- balance-slb bonding
- A source-load-balancing bond mode in AHV that needs no switch configuration; traffic is hashed by source MAC, so a single flow stays on one link.
- LACP bond requirement
- balance-tcp bonding in AHV requires an 802.3ad LACP link aggregation group configured on the physical switch; active-backup and balance-slb do not require it.
- VLAN trunk uplink
- Guest VLANs are 802.1Q tagged by the virtual switch, so the physical uplink must be a trunk port carrying every VLAN used by VM networks.
- virtio NIC driver
- The paravirtualized driver recommended for AHV guests because it delivers the best network and disk throughput; e1000 remains for operating systems lacking virtio support.
- Nutanix Guest Tools
- An in-guest package that enables application-consistent VSS snapshots, Self-Service Restore, file-level restore, and command execution from Prism.
- AHV HA restart
- If an AHV host fails, its VMs restart on surviving hosts; HA reservations pre-reserve capacity so those VMs are guaranteed a place to run.
- VM-VM anti-affinity
- A policy that keeps the listed VMs on different hosts, used to prevent both members of an application pair from failing at the same time.
- AHV maintenance mode
- Entering maintenance mode live-migrates a host's running VMs to other hosts, allowing hardware service on that host with no VM downtime.
- VM clone sources
- An AHV VM can be cloned from a running or powered-off VM, from a snapshot, or from a template; the resulting clone receives independent vDisks.
- AHV overcommit
- AHV allows CPU and memory overcommit, so the sum of configured VM resources can exceed physical host capacity; HA reservation capacity is excluded from overcommit.
- Prism Element vs Central
- Prism Element manages a single cluster; Prism Central manages many registered clusters and adds categories, projects, IAM, Calm, Flow, reports, and Insights.
- Prism Central deployment
- Prism Central runs as a VM on a managed cluster; a single-VM deployment suits small environments and a three-VM deployment provides scale and high availability.
- Categories
- Key-value labels applied to VMs and other objects; protection, recovery, and security policies are attached to categories instead of to individual VMs.
- Prism Central projects
- A project bundles categories, users, quotas, and networks so a team can self-service deploy VMs into an isolated and governed namespace.
- NCC
- Nutanix Cluster Check is the health-check framework that validates services, metadata, disk health, and configuration; it is run from a CVM CLI or Prism.
- allssh command
- A CVM utility that runs a single command on every CVM in the cluster at once, useful for checking one service's status cluster-wide.
- Logbay
- The log collection utility on a CVM, invoked as logbay collect, that packages cluster logs for upload to Nutanix support when opening a case.
- NTP time skew
- Nutanix clusters depend on synchronized time; clock drift between CVMs or hosts causes authentication, certificate, and replication failures.