Alibaba SysOps Administrator flashcards
Alibaba SysOps Administrator · 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.
- ECS instance metadata endpoint
- The link-local address 100.100.100.200 serves instance identity, region, zone, and RAM role credentials to code running on the instance without hard-coded secrets.
- Cloud Assistant
- An ECS agent feature that runs Shell, PowerShell, Bat, or Python commands on instances without SSH and returns invocation results through API or CLI.
- User data first boot
- By default, user data scripts execute only on the initial launch of an ECS instance; repeated execution requires explicit configuration or Cloud Assistant.
- Instance RAM role
- A RAM role attached to an ECS instance that delivers rotating STS credentials via metadata, avoiding long-lived AccessKeys in code or user data.
- VPC CIDR immutability
- A VPC IPv4 CIDR block cannot be modified after creation, so subnet and growth planning must happen before the VPC is provisioned.
- NAT Gateway route
- A custom route table entry sending 0.0.0.0/0 to a NAT Gateway gives private-subnet ECS instances outbound internet without public IPs.
- SLB health check
- A periodic probe of backend port and path; failing backends are removed from rotation, and the check must match the application's actual listener.
- PrivateZone DNS
- A managed private DNS service that resolves custom domain names only inside associated VPCs, useful for internal service discovery.
- OSS storage class minimum duration
- IA objects bill for at least 30 days, Archive 60 days, Cold Archive 180 days, and Deep Cold Archive 180 days regardless of early deletion.
- OSS versioning
- A bucket setting that retains prior object versions and delete markers, enabling recovery from overwrites and accidental deletion at additional storage cost.
- ESSD PL tiers
- ESSD PL0 through PL3 provide increasing IOPS and throughput per disk; PL-X is required for high-IOPS database workloads.
- Snapshot incrementality
- ECS disk snapshots store only changed blocks after the first full snapshot, reducing storage cost and backup time.
- RAM policy evaluation order
- Explicit Deny overrides any Allow; explicit Allow overrides implicit Deny; absence of any Allow results in an implicit Deny.
- STS session duration
- Temporary credentials from STS default to 3600 seconds and can be configured up to 12 hours, after which tokens expire and must be refreshed.
- Security group statefulness
- Security groups track connection state, so return traffic for an allowed inbound connection is automatically permitted without a matching outbound rule.
- ActionTrail audit
- A service that records Alibaba Cloud API management events and can deliver them to OSS or Log Service for compliance and incident investigation.
- CloudMonitor agent metrics
- Memory, disk usage, and process metrics require the CloudMonitor agent on the instance; CPU and network metrics come from the hypervisor without it.
- Alarm silence period
- A configured interval after an alarm fires during which repeated notifications are suppressed, preventing alert storms for the same condition.
- SLS index requirement
- Log Service queries require an index on the Logstore; without an index, full-text and field searches return no results even when data exists.
- CloudMonitor event alarms
- Alarms triggered by system events such as ECS state changes or security group modifications rather than by numeric metric thresholds.
- ROS template sections
- A ROS template contains ROSTemplateFormatVersion, Parameters, Mappings, Conditions, Resources, and Outputs, with Resources declaring Type and Properties.
- RunCommand invocation
- The ECS API and CLI operation that executes a Cloud Assistant command on one or more instances and returns an invocation ID for result polling.
- Alibaba Cloud CLI profile
- A named configuration in ~/.aliyun/config.json holding AccessKey, region, and output format, selectable per command with the --profile flag.
- Terraform OSS backend
- Storing Terraform state in an OSS bucket enables shared, locked state for teams managing Alibaba Cloud resources with the alicloud provider.
- Auto Scaling cooldown
- A period after a scaling activity during which further scaling is suppressed, preventing rapid oscillation when metrics are noisy.
- SLB multi-zone
- A load balancer deployment across a primary and backup zone so the SLB itself survives a zone failure, independent of backend placement.
- RTO and RPO
- Recovery Time Objective is the maximum acceptable downtime; Recovery Point Objective is the maximum acceptable data loss measured in time.
- OSS cross-region replication
- A bucket feature that asynchronously copies objects to a bucket in another region; versioning must be enabled on source and destination.
- SLB 502 diagnosis
- An HTTP 502 from SLB usually indicates a failing health check or a backend refusing connections; verify port, path, and security group rules.
- Disk full remediation
- Identify usage with df and du, clean or rotate logs, then resize the disk and extend the filesystem; snapshots do not reclaim disk space.
- VPC peering prerequisites
- Both VPCs need peering or CEN attachment plus route entries and security group rules allowing traffic; overlapping CIDRs block peering.
- OSS 403 causes
- A 403 from OSS typically stems from bucket policy, RAM policy, or an expired signed URL; check all three before assuming a service outage.