Alibaba ACP Cloud Computing flashcards
Alibaba ACP Cloud Computing · 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.
- Region vs Zone
- A region is a geographic area; zones are isolated datacenters within a region connected by low-latency links. Multi-zone deployment gives fault isolation without cross-region latency.
- Shared responsibility model
- Alibaba Cloud secures physical infrastructure, hypervisor, and managed-service platform; the customer secures guest OS, applications, data, and identity and access configuration.
- VPC scope
- A Virtual Private Cloud is region-scoped, while its vSwitches are zone-scoped. An HA design therefore places vSwitches in at least two zones of one region.
- Resource group
- A logical container for resources within an Alibaba Cloud account, used to group resources for authorization and cost allocation together with tags.
- ECS instance families
- General-purpose (g), compute-optimized (c), memory-optimized (r), plus storage, GPU, and burstable types. Selection is driven by CPU-to-memory ratio and workload profile.
- Subscription vs Pay-As-You-Go
- Subscription is prepaid for a fixed term at a lower unit price; Pay-As-You-Go bills per second or hour with no commitment. Spot instances suit fault-tolerant batch work.
- ESSD disk levels
- Enhanced SSD comes in PL0 through PL3 performance levels; IOPS and throughput increase with PL level and, for some levels, with disk capacity.
- Security group behavior
- A stateful, instance-level virtual firewall. Rules are allow-only, evaluated by priority, and configured separately for inbound and outbound directions.
- Instance metadata service
- An endpoint at 100.100.100.200 reachable from inside an ECS instance that returns instance ID, region, and temporary role credentials to running code.
- EIP vs public IP
- An EIP is a standalone, retainable public IP that can be unbound and rebound to other resources; an instance-assigned public IP is released with the instance.
- NAT Gateway
- A managed VPC gateway that lets private instances without public IPs initiate outbound internet traffic, typically via a 0.0.0.0/0 route entry.
- SLB types
- CLB is the classic layer 4/7 balancer; ALB targets HTTP/HTTPS/gRPC with advanced routing; NLB is a high-performance layer 4 balancer.
- PrivateZone
- A private DNS service that resolves internal domain names only within associated VPCs, enabling stable service names for internal architectures.
- CEN
- Cloud Enterprise Network interconnects many VPCs and on-premises networks through a central routing hub, avoiding a full mesh of VPC peerings.
- OSS storage classes
- Standard, Infrequent Access, Archive, Cold Archive, and Deep Cold Archive. Lower-cost classes trade retrieval time and impose minimum storage durations.
- OSS ZRS vs LRS
- Local Redundant Storage keeps copies within one zone; Zone-Redundant Storage spreads copies across zones in a region for higher durability against zone failure.
- NAS
- A managed shared file system using NFS or SMB that many ECS instances can mount simultaneously, suitable for shared content and lift-and-shift file workloads.
- Hybrid Backup Recovery
- A managed backup service that protects ECS disks and files, NAS, OSS, and on-premises data into backup vaults using scheduled policies and retention rules.
- PolarDB architecture
- A cloud-native database with compute-storage separation; up to 15 read-only nodes share one storage layer, enabling fast cloning and backups.
- RDS read-only instance
- A read-only replica created from an RDS primary that offloads read queries. Applications use read/write splitting to route reads to it.
- RDS point-in-time recovery
- Full and incremental backups plus log backups let you restore an RDS instance to a specific moment within the configured backup retention window.
- Redis cluster edition
- ApsaraDB for Redis cluster edition shards data across multiple nodes for horizontal scale, while standard edition uses a primary/standby pair for high availability.
- RAM role
- An identity assumed by trusted entities such as ECS instances or other accounts, granting temporary STS credentials instead of long-term AccessKeys.
- RAM policy structure
- A JSON document with Effect (Allow/Deny), Action, Resource, and Condition fields; attached to users, groups, or roles to grant scoped permissions.
- WAF
- Web Application Firewall inspects HTTP/HTTPS traffic in front of load balancers to block SQL injection, XSS, and other OWASP web attacks.
- Anti-DDoS scrubbing
- Attack traffic is diverted to a scrubbing center that filters malicious packets and forwards clean traffic to the origin, protecting public-facing IPs.
- Auto Scaling components
- A scaling group defines min/max/desired capacity and zone scope; a scaling configuration is the instance template; rules and alarms trigger scaling.
- Scaling cooldown
- A waiting period after a scaling activity completes during which further scaling actions are suppressed, preventing rapid repeated scale-out and scale-in.
- Lifecycle hook
- An Auto Scaling feature that pauses instance add or remove events so custom scripts can run before the instance joins or leaves the scaling group.
- CloudMonitor alarms
- Metric-based rules that evaluate thresholds over periods and trigger notifications or automated actions such as Auto Scaling when breached.
- SLB 502/504 errors
- These indicate backend health check failure, backend response timeout, or an application error. Verify backend health status, port, and timeout settings.
- Savings plan
- A commitment-based discount where a 1- or 3-year spend commitment lowers the unit price of eligible compute resources compared with pure Pay-As-You-Go.