SAA-C03 Hard Limits & Numbers - Quiz Mode
Master the specific numbers that appear in exam scenarios. This is formatted as a quiz where you can test yourself before revealing the answer.
How to use: Read the question, think of the answer, then expand the section to reveal the correct number and context.
📦 Amazon S3 Limits
▶Q1: What is the minimum number of days before you can transition objects from S3 Standard to S3 Standard-IA?
Answer: 30 days
- Objects must be stored in S3 Standard for at least 30 days before transitioning to Standard-IA
- Minimum storage duration charge for IA storage classes is 30 days (if deleted earlier, you're charged for 30 days)
▶Q2: What is the minimum number of days before you can transition objects from S3 Standard to S3 One Zone-IA?
Answer: 30 days
- Same as Standard-IA: 30-day minimum
- Minimum storage duration charge for One Zone-IA is 30 days
▶Q3: What is the minimum number of days before you can transition to S3 Glacier Flexible Retrieval?
Answer: 90 days
- Objects must be stored for at least 90 days before transitioning to Glacier Flexible Retrieval
- Minimum storage duration charge is 90 days
▶Q4: What is the minimum number of days before you can transition to S3 Glacier Deep Archive?
Answer: 180 days
- Objects must be stored for at least 180 days before transitioning to Deep Archive
- Minimum storage duration charge is 180 days
▶Q5: What is the maximum size of a single S3 object?
Answer: 5 TB (5,000 GB)
- Maximum object size: 5 TB
- For objects larger than 100 MB, use multipart upload (recommended)
- For objects larger than 5 GB, multipart upload is required
▶Q6: What is the maximum size for a single PUT operation in S3?
Answer: 5 GB
- Single PUT operation maximum: 5 GB
- For anything larger, you must use multipart upload
▶Q7: How many S3 buckets can you have per AWS account by default?
Answer: 100 buckets (soft limit, can be increased to 1,000)
- Default limit: 100 buckets per account
- Can request increase to 1,000 via Service Quotas
▶Q8: What is the retrieval time for S3 Glacier Instant Retrieval?
Answer: Milliseconds (same as S3 Standard)
- Retrieval time: Milliseconds
- Use case: Archive data that needs instant access when required
▶Q9: What are the retrieval times for S3 Glacier Flexible Retrieval?
Answer:
- Expedited: 1-5 minutes
- Standard: 3-5 hours
- Bulk: 5-12 hours
▶Q10: What are the retrieval times for S3 Glacier Deep Archive?
Answer:
- Standard: 12 hours
- Bulk: 48 hours
Deep Archive is the cheapest but slowest retrieval option.
⚡ AWS Lambda Limits
▶Q11: What is the maximum execution time (timeout) for a Lambda function?
Answer: 15 minutes (900 seconds)
- Maximum timeout: 15 minutes
- Default timeout: 3 seconds (you must configure higher if needed)
- For longer-running tasks, use ECS/Fargate or Step Functions
▶Q12: What is the maximum memory allocation for a Lambda function?
Answer: 10,240 MB (10 GB)
- Minimum: 128 MB
- Maximum: 10,240 MB (10 GB)
- Memory is allocated in 1 MB increments
- CPU power scales with memory
▶Q13: What is the maximum deployment package size for Lambda (zipped)?
Answer: 50 MB (zipped), 250 MB (unzipped)
- Zipped deployment package: 50 MB
- Unzipped deployment package: 250 MB
- For larger packages, use layers or container images (up to 10 GB)
▶Q14: What is the maximum size for Lambda container images?
Answer: 10 GB
- Container image size: up to 10 GB
- Must be stored in Amazon ECR
▶Q15: What is the maximum number of Lambda layers per function?
Answer: 5 layers
- Maximum layers: 5
- Total unzipped size of layers + deployment package: 250 MB
▶Q16: What is the Lambda concurrent execution limit per region by default?
Answer: 1,000 concurrent executions
- Default: 1,000 concurrent executions per region
- This is a soft limit and can be increased via support ticket
- You can set reserved concurrency for critical functions
▶Q17: What is the maximum payload size for synchronous Lambda invocations?
Answer: 6 MB (request and response)
- Synchronous invocation payload: 6 MB (request and response)
- Asynchronous invocation payload: 256 KB
▶Q18: What is the Lambda environment variable storage limit?
Answer: 4 KB
- Total size of all environment variables: 4 KB
📬 Amazon SQS Limits
▶Q19: What is the default message retention period for SQS?
Answer: 4 days
- Default retention: 4 days
- Minimum: 1 minute
- Maximum: 14 days
▶Q20: What is the maximum message retention period for SQS?
Answer: 14 days
- Maximum retention: 14 days
- After 14 days, messages are automatically deleted
▶Q21: What is the maximum message size in SQS?
Answer: 256 KB
- Maximum message size: 256 KB
- For larger messages, use the SQS Extended Client Library (stores message in S3)
▶Q22: What is the default visibility timeout for SQS messages?
Answer: 30 seconds
- Default visibility timeout: 30 seconds
- Minimum: 0 seconds
- Maximum: 12 hours
▶Q23: What is the maximum visibility timeout for SQS messages?
Answer: 12 hours
- Maximum visibility timeout: 12 hours
- This is how long a message is hidden after being received (before it becomes visible again if not deleted)
▶Q24: What is the delivery delay for SQS FIFO queues?
Answer: 0 seconds (no delay by default)
- FIFO queues: 0-15 minutes delay
- Standard queues: 0-15 minutes delay
- Delay queues postpone delivery of new messages
▶Q25: What is the maximum throughput for SQS FIFO queues?
Answer:
- Without batching: 300 transactions per second (TPS)
- With batching: 3,000 messages per second
FIFO queues have lower throughput than Standard queues.
▶Q26: What is the long polling wait time for SQS?
Answer: 1-20 seconds
- Long polling wait time: 1-20 seconds
- Default (short polling): 0 seconds
- Long polling reduces costs and empty responses
🌊 Amazon Kinesis Limits
▶Q27: What is the default data retention period for Kinesis Data Streams?
Answer: 24 hours
- Default retention: 24 hours
- Can be extended up to 365 days (configurable)
- After retention period, data is deleted
▶Q28: What is the maximum data retention period for Kinesis Data Streams?
Answer: 365 days (1 year)
- Maximum retention: 365 days
- Enables replay of data within this window
▶Q29: What is the maximum data blob size for a single Kinesis Data Streams record?
Answer: 1 MB
- Maximum record size: 1 MB
- For larger data, split into multiple records
▶Q30: What is the write capacity of a single Kinesis shard?
Answer: 1 MB/sec or 1,000 records/sec
- Write capacity per shard: 1 MB/sec OR 1,000 records/sec (whichever is hit first)
▶Q31: What is the read capacity of a single Kinesis shard?
Answer: 2 MB/sec
- Read capacity per shard: 2 MB/sec
- This applies to all consumers combined (classic mode)
- Enhanced fan-out provides 2 MB/sec per consumer
▶Q32: What is the maximum number of shards per Kinesis Data Stream in a region?
Answer: 500 (default soft limit)
- Default limit: 500 shards per region
- This is a soft limit and can be increased
▶Q33: What is the minimum buffer interval for Kinesis Data Firehose delivery to S3?
Answer: 60 seconds
- Minimum buffer interval: 60 seconds
- Firehose is near real-time (not real-time like Data Streams)
- Delivers data based on buffer size or time interval (whichever is hit first)
▶Q34: What is the maximum buffer size for Kinesis Data Firehose delivery to S3?
Answer: 128 MB
- Maximum buffer size: 128 MB
- Firehose delivers when buffer size OR time interval is reached
💾 Amazon EBS Limits
▶Q35: What is the maximum size of a single EBS volume?
Answer: 64 TiB (tebibytes)
- Maximum volume size: 64 TiB for all volume types
- 1 TiB = 1,024 GiB
▶Q36: What is the maximum IOPS for an EBS io2 Block Express volume?
Answer: 256,000 IOPS
- io2 Block Express: up to 256,000 IOPS
- io2 (non-Block Express): up to 64,000 IOPS
- io1: up to 64,000 IOPS
▶Q37: What is the maximum throughput for an EBS io2 Block Express volume?
Answer: 4,000 MiB/s
- io2 Block Express: up to 4,000 MiB/s
- io2/io1: up to 1,000 MiB/s
▶Q38: What is the maximum IOPS for an EBS gp3 (General Purpose SSD) volume?
Answer: 16,000 IOPS
- gp3: up to 16,000 IOPS
- gp2: up to 16,000 IOPS (but based on volume size: 3 IOPS per GiB)
▶Q39: What is the maximum throughput for an EBS gp3 volume?
Answer: 1,000 MiB/s
- gp3: up to 1,000 MiB/s (independently configurable)
- gp2: up to 250 MiB/s
▶Q40: How many EBS volumes can you attach to a single EC2 instance?
Answer: Most instances support up to 28 volumes
- Depends on instance type
- Most instances: up to 28 volumes
- Nitro-based instances have higher limits
▶Q41: How many EC2 instances can attach to a single EBS volume (with Multi-Attach)?
Answer: Up to 16 instances (io1/io2 Provisioned IOPS only)
- io1/io2 Multi-Attach: up to 16 instances
- Only available for Provisioned IOPS volumes
- All instances must be in the same AZ
🗄️ Amazon RDS & Aurora Limits
▶Q42: What is the maximum storage capacity for RDS instances (for most engines)?
Answer: 64 TiB
- Maximum RDS storage: 64 TiB (for MySQL, PostgreSQL, MariaDB, Oracle, SQL Server)
- Aurora: Auto-scales up to 128 TiB
▶Q43: What is the maximum storage capacity for Amazon Aurora?
Answer: 128 TiB (auto-scaling)
- Aurora automatically scales storage from 10 GB to 128 TiB
- No need to provision storage upfront
▶Q44: How many read replicas can you create for standard RDS instances?
Answer: 5 read replicas
- Standard RDS: up to 5 read replicas
- Aurora: up to 15 read replicas
▶Q45: How many read replicas can you create for Amazon Aurora?
Answer: 15 read replicas
- Aurora: up to 15 read replicas
- Aurora replicas have low-latency promotion (< 30 seconds)
▶Q46: What is the maximum automated backup retention period for RDS?
Answer: 35 days
- Maximum automated backup retention: 35 days
- Point-in-time recovery (PITR) within this window
- Manual snapshots can be retained indefinitely
▶Q47: What is Aurora's typical failover time?
Answer: < 30 seconds
- Aurora failover: typically < 30 seconds
- Standard RDS Multi-AZ failover: 1-2 minutes
▶Q48: What is the maximum number of Aurora Global Database secondary regions?
Answer: 5 secondary regions
- Aurora Global Database: up to 5 secondary regions
- Replication lag: typically < 1 second
🔄 Amazon DynamoDB Limits
▶Q49: What is the maximum item size in DynamoDB?
Answer: 400 KB
- Maximum item size: 400 KB
- Includes both attribute names and values
▶Q50: How many Global Secondary Indexes (GSI) can you create per DynamoDB table?
Answer: 20 GSIs
- Maximum GSIs: 20 per table
- GSIs can be created/deleted after table creation
▶Q51: How many Local Secondary Indexes (LSI) can you create per DynamoDB table?
Answer: 5 LSIs
- Maximum LSIs: 5 per table
- LSIs must be created at table creation time (cannot be added later)
▶Q52: What is the maximum number of partition keys you can specify in a DynamoDB BatchGetItem request?
Answer: 100 items
- BatchGetItem: up to 100 items
- BatchWriteItem: up to 25 items
▶Q53: What is the maximum number of items in a DynamoDB BatchWriteItem request?
Answer: 25 items
- BatchWriteItem: up to 25 items
- Total request size: up to 16 MB
▶Q54: What is DynamoDB's point-in-time recovery (PITR) retention period?
Answer: 35 days
- PITR retention: 35 days
- Allows you to restore to any point in time within the last 35 days
🌐 VPC & Networking Limits
▶Q55: How many VPCs can you create per region by default?
Answer: 5 VPCs per region
- Default limit: 5 VPCs per region
- Soft limit (can be increased)
▶Q56: How many subnets can you create per VPC?
Answer: 200 subnets
- Maximum subnets per VPC: 200
- Soft limit (can be increased)
▶Q57: How many IPv4 CIDR blocks can you associate with a VPC?
Answer: 5 CIDR blocks
- Primary CIDR block: 1
- Additional secondary CIDR blocks: 4
- Total: 5 CIDR blocks per VPC
▶Q58: How many Elastic IP addresses can you have per region by default?
Answer: 5 Elastic IPs
- Default limit: 5 Elastic IPs per region
- Soft limit (can be increased)
▶Q59: How many rules can you have per security group?
Answer: 60 inbound + 60 outbound = 120 total
- Inbound rules: 60
- Outbound rules: 60
▶Q60: How many security groups can you attach to a single ENI (network interface)?
Answer: 5 security groups
- Maximum security groups per ENI: 5
▶Q61: How many rules can you have per Network ACL?
Answer: 20 inbound + 20 outbound (default)
- Default: 20 inbound + 20 outbound
- Can be increased to 40 each (soft limit)
▶Q62: What is the maximum bandwidth for a NAT Gateway?
Answer: 45 Gbps (auto-scaling)
- NAT Gateway: up to 45 Gbps (scales automatically)
- NAT Instance: depends on instance type
▶Q63: How many VPC peering connections can you have per VPC?
Answer: 125 active peering connections
- Maximum peering connections per VPC: 125
▶Q64: How many VPN connections can you have per VPC?
Answer: 10 VPN connections (default soft limit)
- Default: 10 VPN connections per VPC
- Can be increased
⚖️ Elastic Load Balancing Limits
▶Q65: How many targets can you register with a single Application Load Balancer?
Answer: 1,000 targets
- ALB: up to 1,000 targets per load balancer
- This includes all target groups combined
▶Q66: How many listeners can you have per Application Load Balancer?
Answer: 50 listeners
- ALB/NLB: up to 50 listeners per load balancer
▶Q67: How many rules can you have per ALB listener?
Answer: 100 rules (default)
- Default: 100 rules per listener
- Can be increased to 1,000 (soft limit)
▶Q68: How many certificates can you install on an ALB?
Answer: 25 certificates (default)
- Default: 25 certificates per load balancer
- Can be increased to 100 (soft limit)
▶Q69: What is the idle timeout for an Application Load Balancer connection?
Answer: 60 seconds (default)
- Default idle timeout: 60 seconds
- Can be configured from 1 to 4,000 seconds
🚗 Amazon EC2 Limits
▶Q70: How many On-Demand instances can you run per region by default (vCPU limit)?
Answer: Varies by instance family (vCPU-based limits)
- EC2 uses vCPU limits per instance family
- Standard instances: typically 1,152 vCPUs per region (default)
- Limits vary by account age and usage
▶Q71: How many Elastic Network Interfaces (ENIs) can you attach to a single EC2 instance?
Answer: Varies by instance type
- Depends on instance size
- Example: t3.micro = 2 ENIs, m5.large = 3 ENIs, m5.24xlarge = 15 ENIs
▶Q72: How many tags can you assign to an EC2 instance?
Answer: 50 tags
- Maximum tags per resource: 50
▶Q73: What is the maximum lifetime of a Spot Instance?
Answer: No maximum (runs until interrupted or terminated)
- Spot Instances can run indefinitely
- Terminated by AWS when capacity is needed or price exceeds your bid
🔐 IAM Limits
▶Q74: How many IAM users can you create per AWS account?
Answer: 5,000 IAM users
- Maximum IAM users: 5,000 per account
- For more users, use IAM Identity Center (SSO) or federation
▶Q75: How many IAM groups can you create per AWS account?
Answer: 300 groups
- Maximum IAM groups: 300 per account
▶Q76: How many IAM groups can a single user belong to?
Answer: 10 groups
- Maximum groups per user: 10
▶Q77: How many managed policies can you attach to an IAM user, group, or role?
Answer: 10 managed policies
- Maximum managed policies per principal: 10
- You can also have 1 inline policy per principal
▶Q78: What is the maximum size of an IAM policy document?
Answer: 6,144 characters (for managed policies)
- Managed policies: 6,144 characters
- Inline policies (user): 2,048 characters
- Inline policies (role, group): 10,240 characters
▶Q79: How many IAM roles can you create per AWS account?
Answer: 1,000 roles
- Maximum IAM roles: 1,000 per account
▶Q80: What is the maximum session duration for an IAM role assumption?
Answer: 12 hours
- Maximum session duration: 12 hours
- Default: 1 hour (configurable)
📧 Amazon SNS Limits
▶Q81: What is the maximum message size for Amazon SNS?
Answer: 256 KB
- Maximum message size: 256 KB
- For larger messages, store in S3 and send S3 link
▶Q82: How many subscriptions can you have per SNS topic?
Answer: 12,500,000 subscriptions (default)
- Default: 12.5 million subscriptions per topic
▶Q83: How many topics can you create per AWS account?
Answer: 100,000 topics
- Maximum SNS topics: 100,000 per account
📊 Amazon CloudWatch Limits
▶Q84: How long are CloudWatch metrics stored?
Answer: 15 months
- Metric retention: 15 months
- You can retrieve metrics older than 2 weeks using GetMetricStatistics API
▶Q85: What is the maximum number of dimensions you can include in a CloudWatch metric?
Answer: 10 dimensions
- Maximum dimensions per metric: 10
▶Q86: How many alarms can you create per region?
Answer: 5,000 alarms (default soft limit)
- Default: 5,000 alarms per region
- Can be increased
▶Q87: What is the minimum resolution for CloudWatch custom metrics?
Answer: 1 second (high-resolution metrics)
- Standard resolution: 1 minute
- High-resolution: 1 second
- High-resolution metrics retained for 3 hours at 1-second resolution
🏗️ AWS CloudFormation Limits
▶Q88: How many stacks can you create per region?
Answer: 2,000 stacks (default)
- Default: 2,000 stacks per region
- Includes active and deleted stacks (within 90 days)
▶Q89: What is the maximum template body size for CloudFormation?
Answer: 51,200 bytes (when passed directly), 1 MB (when uploaded to S3)
- Direct upload: 51,200 bytes
- S3 upload: 1 MB
▶Q90: How many parameters can you declare in a CloudFormation template?
Answer: 200 parameters
- Maximum parameters: 200 per template
▶Q91: How many outputs can you declare in a CloudFormation template?
Answer: 200 outputs
- Maximum outputs: 200 per template
🌍 Amazon Route 53 Limits
▶Q92: How many hosted zones can you create per AWS account?
Answer: 500 hosted zones
- Default limit: 500 hosted zones per account
- Soft limit (can be increased)
▶Q93: How many records can you create per hosted zone?
Answer: 10,000 records
- Maximum records per hosted zone: 10,000
- Soft limit (can be increased)
▶Q94: How many health checks can you create per AWS account?
Answer: 200 health checks (default)
- Default: 200 health checks
- Soft limit (can be increased to 1,000)
▶Q95: What is the TTL (Time to Live) range for Route 53 DNS records?
Answer: 0 seconds to 2,147,483,647 seconds
- Minimum TTL: 0 seconds (not recommended)
- Maximum TTL: 2,147,483,647 seconds (~68 years)
- Common TTLs: 60-300 seconds for frequently changing records
🔄 AWS Step Functions Limits
▶Q96: What is the maximum execution time for a Step Functions Standard workflow?
Answer: 1 year
- Standard workflow: up to 1 year
- Express workflow: up to 5 minutes
▶Q97: What is the maximum execution time for a Step Functions Express workflow?
Answer: 5 minutes
- Express workflow: up to 5 minutes
- Express workflows are designed for high-volume, short-duration tasks
▶Q98: What is the maximum execution history size for Step Functions?
Answer: 25,000 events
- Standard workflow: 25,000 events in execution history
🗂️ AWS Storage Gateway Limits
▶Q99: What is the maximum size of a file for File Gateway?
Answer: 5 TB
- Maximum file size: 5 TB (same as S3 object limit)
▶Q100: What is the recommended cache size for Storage Gateway (File/Volume Gateway)?
Answer: At least 150 GB, recommended to be 20% of your data set
- Minimum cache: 150 GB
- Recommended: 20% of total dataset size
- Maximum cache volume size: 64 TiB
▶Q101: How many volumes can you create per Volume Gateway?
Answer: 32 volumes (cached mode) or 32 volumes (stored mode)
- Maximum volumes: 32 per gateway
- Maximum size per volume: 32 TiB (cached), 16 TiB (stored)
▶Q102: What is the maximum size for a Storage Gateway cached volume?
Answer: 32 TiB
- Cached volume: up to 32 TiB
- Stored volume: up to 16 TiB
💼 Amazon ECS & EKS Limits
▶Q103: How many tasks can you run per ECS cluster?
Answer: 10,000 tasks (default soft limit)
- Default: 10,000 tasks per cluster
- Can be increased
▶Q104: How many services can you create per ECS cluster?
Answer: 5,000 services
- Maximum services: 5,000 per cluster
▶Q105: How many container instances can you have per ECS cluster?
Answer: 10,000 instances (default soft limit)
- Default: 10,000 container instances per cluster
▶Q106: What is the maximum task definition size in ECS?
Answer: 64 KiB
- Maximum task definition size: 64 KiB (for the JSON document)
🚀 Amazon CloudFront Limits
▶Q107: How many distributions can you create per AWS account?
Answer: 200 distributions (default)
- Default: 200 distributions per account
- Soft limit (can be increased)
▶Q108: How many alternate domain names (CNAMEs) can you have per CloudFront distribution?
Answer: 100 CNAMEs
- Maximum CNAMEs: 100 per distribution
▶Q109: How many cache behaviors can you have per CloudFront distribution?
Answer: 25 cache behaviors (default)
- Default: 25 cache behaviors per distribution
- Can be increased to 1,000
▶Q110: What is the maximum file size CloudFront can serve?
Answer: 30 GB (for HTTP methods), 20 GB (for PUT/POST/PATCH)
- GET/HEAD: 30 GB
- PUT/POST/PATCH: 20 GB
▶Q111: What is the default TTL for CloudFront cached objects?
Answer: 24 hours (86,400 seconds)
- Default TTL: 24 hours
- Can be configured (minimum 0 seconds, maximum 1 year)
🔑 AWS Secrets Manager Limits
▶Q112: What is the maximum size of a secret in AWS Secrets Manager?
Answer: 65,536 bytes (64 KB)
- Maximum secret size: 64 KB
▶Q113: How many versions can a secret have in Secrets Manager?
Answer: ~100 versions (soft limit)
- Approximately 100 versions per secret
- Older versions are automatically removed
🎯 Amazon API Gateway Limits
▶Q114: What is the maximum integration timeout for API Gateway?
Answer: 29 seconds
- Maximum timeout: 29 seconds
- This applies to all integrations (Lambda, HTTP, etc.)
▶Q115: What is the maximum payload size for API Gateway?
Answer: 10 MB
- Maximum request/response payload: 10 MB
▶Q116: What is the default throttle limit for API Gateway (requests per second)?
Answer: 10,000 requests per second (steady-state), 5,000 burst
- Steady-state: 10,000 RPS
- Burst: 5,000 requests
- These are regional limits and can be increased
▶Q117: How many stages can you have per API in API Gateway?
Answer: 10 stages
- Maximum stages: 10 per API
- Common stages: dev, test, prod
🎬 Amazon EventBridge Limits
▶Q118: What is the maximum size of an EventBridge event?
Answer: 256 KB
- Maximum event size: 256 KB
▶Q119: How many rules can you create per event bus in EventBridge?
Answer: 300 rules (default)
- Default: 300 rules per event bus
- Can be increased
▶Q120: How many targets can you have per EventBridge rule?
Answer: 5 targets
- Maximum targets: 5 per rule
🧠 Amazon SageMaker Limits
▶Q121: What is the maximum runtime for a SageMaker training job?
Answer: 28 days
- Maximum training job duration: 28 days
▶Q122: What is the maximum size of a SageMaker model artifact?
Answer: 5 GB (or unlimited if using S3)
- Compressed model: 5 GB (if uploaded directly)
- For larger models, use S3
🏷️ AWS Resource Tags Limits
▶Q123: How many tags can you assign to most AWS resources?
Answer: 50 tags
- Maximum tags: 50 per resource
- Applies to most AWS resources (EC2, S3, RDS, etc.)
▶Q124: What is the maximum key length for a resource tag?
Answer: 128 characters
- Tag key: up to 128 characters
- Tag value: up to 256 characters
🌐 AWS Direct Connect Limits
▶Q125: What are the available port speeds for AWS Direct Connect?
Answer: 1 Gbps, 10 Gbps, or 100 Gbps (dedicated connections)
- Dedicated connections: 1 Gbps, 10 Gbps, 100 Gbps
- Hosted connections: 50 Mbps, 100 Mbps, 200 Mbps, 300 Mbps, 400 Mbps, 500 Mbps, 1 Gbps, 2 Gbps, 5 Gbps, 10 Gbps
▶Q126: How many virtual interfaces (VIFs) can you create per Direct Connect connection?
Answer: 50 virtual interfaces (default)
- Default: 50 VIFs per connection
- Can be increased
🏢 AWS Organizations Limits
▶Q127: How many AWS accounts can you have in an AWS Organization?
Answer: 10 accounts (default), can be increased to thousands
- Default: 10 accounts
- Can request increase via support (typically approved for hundreds/thousands)
▶Q128: How many levels of Organizational Units (OUs) can you create?
Answer: 5 levels deep
- Maximum OU hierarchy depth: 5 levels
- Root does not count as a level
▶Q129: How many Service Control Policies (SCPs) can you attach to an OU or account?
Answer: 5 SCPs
- Maximum SCPs per OU/account: 5
🔍 Amazon Elasticsearch/OpenSearch Limits
▶Q130: What is the maximum number of indices per Amazon OpenSearch domain?
Answer: 1,000 indices (default soft limit)
- Default: 1,000 indices
- Can be increased
▶Q131: What is the maximum domain size for Amazon OpenSearch?
Answer: 3 PB (petabytes)
- Maximum domain storage: 3 PB
🧩 AWS Batch Limits
▶Q132: What is the maximum number of job queues in AWS Batch?
Answer: 50 job queues
- Maximum job queues: 50 per region
▶Q133: What is the maximum number of compute environments in AWS Batch?
Answer: 50 compute environments
- Maximum compute environments: 50 per region
📦 Amazon ECR Limits
▶Q134: What is the maximum size of a container image in Amazon ECR?
Answer: 10 GB per layer, no limit on total image size
- Maximum layer size: 10 GB
- No limit on total image size (but practical limits apply)
▶Q135: How many repositories can you create in Amazon ECR?
Answer: 10,000 repositories (default soft limit)
- Default: 10,000 repositories per region
🔐 AWS KMS Limits
▶Q136: What is the maximum request rate for AWS KMS API calls?
Answer: Varies by API and region (5,500 - 30,000 requests per second)
- Shared quota (most APIs): 5,500 RPS (varies by region)
- Some operations (like GenerateDataKey): up to 30,000 RPS
- These are soft limits
▶Q137: What is the maximum size of data you can encrypt directly with KMS?
Answer: 4 KB
- Direct encryption: up to 4 KB
- For larger data, use envelope encryption (GenerateDataKey)
▶Q138: How many grants can you have per KMS key?
Answer: 50,000 grants (default soft limit)
- Default: 50,000 grants per key
🌊 Amazon MSK (Managed Streaming for Kafka) Limits
▶Q139: What is the maximum message size for Amazon MSK?
Answer: 1 MB (default), can be increased to 10 MB
- Default: 1 MB
- Configurable up to 10 MB
▶Q140: How many brokers can you have per MSK cluster?
Answer: No hard limit (limited by instance capacity)
- No specific limit (depends on instance types and data volume)
- Typically start with 3 brokers (one per AZ for HA)
🎥 Amazon Kinesis Video Streams Limits
▶Q141: What is the default data retention period for Kinesis Video Streams?
Answer: 24 hours (default), can be increased to 7 years
- Default: 24 hours
- Can be configured up to 87,600 hours (10 years)
▶Q142: What is the maximum number of concurrent connections per Kinesis Video Stream?
Answer: 5 concurrent connections
- Maximum readers: 5 concurrent connections per stream
🌍 AWS Global Accelerator Limits
▶Q143: How many static IP addresses does AWS Global Accelerator provide?
Answer: 2 static Anycast IPs
- Global Accelerator provides 2 static Anycast IPs
- These IPs are fixed entry points to your application
▶Q144: How many endpoints can you have per endpoint group in Global Accelerator?
Answer: 10 endpoints
- Maximum endpoints: 10 per endpoint group
🧪 AWS Systems Manager Limits
▶Q145: What is the maximum size of a parameter in Systems Manager Parameter Store (Standard tier)?
Answer: 4 KB
- Standard tier: 4 KB
- Advanced tier: 8 KB
▶Q146: How many parameters can you store in Parameter Store (Standard tier)?
Answer: 10,000 parameters
- Standard tier: 10,000 parameters (free)
- Advanced tier: 100,000 parameters (charged)
📦 AWS Transfer Family Limits
▶Q147: What is the maximum file size for AWS Transfer Family?
Answer: 5 TB (same as S3)
- Maximum file size: 5 TB
- Same as underlying S3 object limit
▶Q148: How many users can you have per Transfer Family server?
Answer: 10,000 users (default soft limit)
- Default: 10,000 users per server
🔄 AWS DataSync Limits
▶Q149: What is the maximum number of files DataSync can transfer per task?
Answer: 50 million files
- Maximum files: 50 million per task
▶Q150: What is the maximum transfer rate for DataSync?
Answer: 10 Gbps per task
- Maximum throughput: 10 Gbps per task
- Can run multiple tasks in parallel for higher throughput
📊 Summary: Critical Numbers to Memorize
Here are the most exam-critical numbers you should commit to memory:
Storage & Database
- S3 max object size: 5 TB
- S3 Lifecycle transitions: 30 days (IA), 90 days (Glacier), 180 days (Deep Archive)
- EBS max volume size: 64 TiB
- RDS read replicas: 5 (standard), 15 (Aurora)
- RDS backup retention: 35 days max
- DynamoDB max item size: 400 KB
Compute
- Lambda max execution time: 15 minutes
- Lambda max memory: 10 GB
- Lambda concurrent executions: 1,000 (default)
Messaging & Streaming
- SQS message retention: 14 days max, 4 days default
- SQS message size: 256 KB max
- SQS visibility timeout: 12 hours max
- Kinesis Data Streams retention: 24 hours default, 365 days max
- Kinesis shard write: 1 MB/sec or 1,000 records/sec
- Kinesis shard read: 2 MB/sec
Networking
- NAT Gateway bandwidth: 45 Gbps (auto-scaling)
- VPC peering connections: 125 per VPC
- Security groups per ENI: 5
- Direct Connect port speeds: 1/10/100 Gbps
Other Critical Numbers
- API Gateway timeout: 29 seconds
- IAM users per account: 5,000
- Step Functions (Standard): 1 year max
- Step Functions (Express): 5 minutes max
- CloudWatch metrics retention: 15 months
🎓 Exam Strategy Tips
- Pattern Recognition: If you see "compliance" or "long-term retention" → think Deep Archive (180 days minimum)
- 30-90-180 Rule: Remember S3 storage class transition days (30 for IA, 90 for Glacier, 180 for Deep Archive)
- Lambda = 15 minutes: Any scenario requiring longer = use ECS/Fargate/Step Functions
- SQS 14 days: If messages need to be retained longer, use Kinesis Data Streams (365 days)
- Kinesis shard math: Each shard = 1 MB write, 2 MB read. Calculate shards needed based on throughput requirements.
📝 Practice Questions
Test yourself by asking:
- "If my application generates 5,000 records per second at 1 KB each, how many Kinesis shards do I need?"
- "Can I use Lambda for a video processing job that takes 20 minutes?" (No → ECS/Fargate)
- "My S3 lifecycle policy transitions after 60 days to Glacier. What's wrong?" (Need 90 days minimum)
Good luck with your SAA-C03 exam! 🚀