Cloudamize Data Collection Methodologies – Agent vs. Agentless

Overview

To provide precise analytics for cloud right-sizing, cost projections, and application migration plans, the Cloudamize platform collects deep system and application-level performance metrics. Cloudamize offers two primary methods for this data collection: Agent-Based (Recommended) and the Agentless Data Collector (ADC).

Choosing the appropriate methodology depends on your organization's security posture, deployment flexibility, and the level of data resolution required for your migration planner.

Architecture & Technical Comparison

The table below outlines the architecture, capabilities, and fundamental core differences between the two options.

Feature / Metric	Cloudamize Agent-Based (Recommended)	Cloudamize Agentless Data Collector (ADC)
Deployment Model	Installed directly on every target machine (Windows/Linux).	Installed on one dedicated Windows Server / VM per subnet.
Scaling Capacity	Boundless (independent agent-based scale).	Monitors up to 500 hosts per collector by default.
Target Communications	Initiated natively on the local OS.	Uses WMI/RPC (Windows) and SSH (Linux).
Firewall Requirements	Outbound Only: Port TCP 443 (SSL) to Cloudamize servers. Can route via an internal authenticated proxy.	Inbound & Outbound: Requires opening multiple inbound ports per endpoint for the collector to pull data.
Performance Sampling	High-resolution: Every 30 seconds.	Lower resolution: Every 5 minutes.
Application Dependencies	Captured seamlessly (including short-lived/bursty connections and UDP traffic).	Limited. Misses bursty/short-lived connections (80% of server connections) and cannot read the UDP protocol.
Performance Throttling	Built-in smart Watchdog process to dynamically cap overhead if CPU exceeds 2%.	None. No performance-controlling mechanism on the monitored host.
Migration Enablement	Allows automated installation of CSP migration agents (e.g., AWS MGN agent) from the Cloudamize platform.	Discovery and assessment only; does not support an automated migration tool.
Encryption Standard	TLS v1.2 / TLS v1.3	TLS v1.2 / TLS v1.3

Detailed Analysis: Advantages & Disadvantages

1. Cloudamize Agent-Based Data Collector

The Agent-Based framework is the vendor-recommended approach due to its granularity and lightweight footprinting.

Advantages:
- Precision Right-Sizing: Collects granular analytics every 30 seconds, catching short-term utilization peaks and ensuring highly accurate cloud configuration recommendations.
- Full Dependency Mapping: Successfully reads both TCP and UDP protocols, mapping 100% of application network connections (even bursty or short-lived ones).
- Simplified Egress Security: Only requires a single outbound connection (Port 443) or an existing corporate proxy. No internal inbound firewall rules need to be altered.
- Fail-safe Resource Limits: Employs a "Watchdog Agent" alongside the collection agent. If local server resource utilization exceeds a specific threshold (e.g., CPU usage above 2%), the agent automatically goes to sleep to minimize host performance impact.
- Downstream Integration: Streamlines migration by allowing users to natively push cloud migration software (like AWS Migration Evaluator/MGN) directly through the console later on.
Disadvantages:
- Management Overhead: Requires installation access across all global servers, which can mandate complex software distribution cycles (SCCM, Ansible, etc.) in massive environments.
- OS Prerequisites: Requires .NET Framework 3.5+ on Windows endpoints and specific Linux kernel configurations with root privileges for mapping network dependencies (PID-to-network).

2. Cloudamize Agentless Data Collector (ADC)

The Agentless approach relies on a central Windows appliance scanning the network subnet to gather infrastructure performance metrics remotely.

Advantages:
- No Target Footprint: Ideal for highly restricted or compliant corporate infrastructures that strictly prohibit installing third-party agent software on production nodes.
- Centralized Administration: Installed and maintained on a single machine. Updates and configurations are handled globally by a single server.
- Internal Egress Isolation: Data from the entire infrastructure is aggregated locally and exits the corporate network through a single secure portal (the ADC machine).
Disadvantages:
- Blind Spots in Dependencies: Because it monitors at a 5-minute polling interval, it may miss bursty or short-lived connections. This means critical application links may be missing from your Migration Move Groups.
- Internal Security Changes: Requires setting up widespread inbound firewall access over WMI, RPC, and SSH across subnets and VLANs, which often complicates security approvals.
- Credential Requirements: The central collector must be configured with broad domain administration privileges or SSH root credentials to scrape data from target machines.
- SQL Limitations: While it can gather basic Microsoft SQL metrics, it relies heavily on remote PowerShell configurations.

Complete Feature & Data Collection Comparison Matrix

1. System Inventory & Infrastructure Telemetry

This table lists all system information metadata and performance-tracking points as detailed in the Data Collected by Agent and Agentless Overview documentation.

Category	Telemetry / Data Point	Captured by Agent	Captured by Agentless (ADC)	Collection Mechanism & Impact
Time Trackers	Last shutdown time	✔	✔	Captured via local registry / WMI event logs.
	Last boot-up time	✔	✔	Captures host machine uptime intervals.
Storage Hardware	Disk drive metadata	✔	✔	Drive Name, Interface type, Manufacturer, Serial number, and Description.
Compute Hardware	Processor Type	✔	✔	e.g., `"Intel(R) Xeon(R) CPU X5482 @ 3.20GHz"`
	Processor Attributes	✔	✔	Clock rate, processor family, and physical core counts.
	Memory Capacity	✔	✔	Total provisioned physical RAM allocations.
Operating System	OS Build Details	✔	✔	Exact text e.g., `Microsoft Windows [Version 6.2.9200]`.
Identity & Scope	System Identification	✔	✔	DNS host name, System domain name, System name, Workgroup string, and Domain-membership flag (`True`/`False`).
	Virtualization Identity	✔	✔	VM Name (if virtualized under VMware vCenter or Hyper-V layers).
Networking	Network Interfaces (NIC)	✔	✔	MAC address(es), IP address(es), and NIC types (e.g., `Intel(R) Gigabit Connection`).
System Status	Overall Host Status	✔	✔	Availability state signals.
High Frequency Performance	CPU Usage Profile	✔ (30s)	✔ (5m)	Agentless Warning: Misses short-term operational peaks due to 5-minute averaging gaps.
	Memory Usage Profile	✔ (30s)	✔ (5m)	Tracks memory commitments and page file activity.
	Disk Operations (IOPS)	✔ (30s)	✔ (5m)	Measures read/write storage strain factors.
	Network Throughput	✔ (30s)	✔ (5m)	Assesses inbound/outbound bandwidth volumes.
	Cache Pool Usage	✔ (30s)	✔ (5m)	Tracks internal system layer buffer caches.

2. Process & Application Connectivity Telemetry

Category	Telemetry / Data Point	Captured by Agent	Captured by Agentless (ADC)	Technical Constraints & Impact
Process Metadata	Executable Filename	✔	✔	Identifies application boundaries (e.g., `sqlservr.exe`, `apache2`).
	Software Vendor	✔	✔	Resolves corporate publisher certificates.
	Application Information	✔	✔	Pulls metadata such as Product name, description, and source URL.
Process Metrics	Isolated Process Usage	✔	✔	Segregates CPU, Memory, Disk, and Network usage per PID.
	Performance Counters	✔	No	Integrates application-specific system performance counters.
Software Inventory	Installed Program List	✔	✔	Exhaustive historic dump of installed apps from system controls.
Network Maps	Active TCP Mappings	✔	✔	Agentless: Misses short-lived/bursty connections (up to 80% of network dependencies).
	Active UDP Connections	✔	✔	Stateless UDP streams cannot be traced by remote polling.
	Directional Data Flow	✔	✔	Identifies if the traffic path is inbound, outbound, or bi-directional.
	Connection Throughput	✔	✔	Calculates exact transfer speeds (KB/s) per application link.
	Process-to-Port Linkage	✔	✔	Correlates which specific process opened a specific network link.

3. Microsoft SQL Server Telemetry (Basic & Advanced)

Both data collection methods use a PowerShell script engine to poll SQL workloads.

Agent Prerequisites: Natively captures Basic data. Advanced data requires PowerShell 5.1+.
Agentless Prerequisites: Basic data requires PowerShell 2.0 or later on the target node. Advanced data requires PowerShell 5.1+ on the target node, and the remote user credential must explicitly hold sysadmin/read privileges on the instance.

Basic SQL Data Comparison

Telemetry Data Point	Captured by Agent	Captured by Agentless (ADC)	Core Context / Sample Output
SQL Server Instance ID	✔	✔	e.g., `MSSQL11.MSSQLSERVER`
SQL Server Instance Name	✔	✔	e.g., `MSSQLSERVER` or custom named instances.
SQL Server Edition	✔	✔	e.g., `Standard Edition`, `Enterprise Edition`
SQL Server Version	✔	✔	e.g., Engine Build String `9.4.5000.00`
SQL Server Patch Level	✔	✔	Tracks Cumulative Update (CU) configurations.
SQL Server Service Pack	✔	✔	e.g., `Service Pack 4`
SQL Server Bin Path	✔	✔	Executable root directory pathing string.
SQL Server Data Path	✔	✔	Default data cluster directory tracking string.
SQL Server Error Log File	✔	✔	Full string location pointing directly to the system `ERRORLOG`.
Database Names & Status	✔	✔	Full inventory list (e.g., `master`, `model`, `UserDB_1`).
Database Size	✔	✔	Captures active size metrics calculated in Megabytes (MB).
Database File Locations	✔	✔	Location strings mapping `.mdf` data and `.ldf` log files.
Core Licenses Text	✔	✔	Required documentation data for licensing analysis.
Cluster Name	✔	✔	Failover Cluster Instance network identifying string.
Cluster State	✔	✔	Verifies if the node is part of a failover cluster layer.
SQL Server Node Status	✔	✔	Flags if the evaluated instance node is Active or Passive.
SQL Cluster Owner Host	✔	✔	Identifies the physical host currently controlling the cluster group.
SQL Server Feature List	✔	✔	Identifies internal sub-component features deployed locally.
SQL Server Product Info	✔	✔	Base engine SKU product validation streams.
SQL Server HADR Usage	✔	✔	Verifies if Availability Group frameworks are active.
SQL FileStream Level	✔	✔	Configuration level number for local FileStream parameters.
Is FullText Installed	✔	✔	True/False confirmation of Full-Text Search integration.
Is PolyBase Installed	✔	✔	True/False status of PolyBase query distribution configuration.
Collation Content	✔	✔	Default language/sorting database collation logic strings.

Advanced SQL Data Comparison (Required for Cloud Optimization & OLA)

Telemetry Data Point	Captured by Agent	Captured by Agentless (ADC)	Core Context / Sample Output
Cluster Quorum Type	✔	✔	Topology validation (e.g., Node Majority, Witness Disk).
PerfMon Mode	✔	✔	Captures active counter polling logic states.
Availability Group Name	✔	✔	Name string of AlwaysOn Availability Groups (AG).
AG Primary Replica Server	✔	✔	Pinpoints the active write-replica server destination.
AG Local Replica Role	✔	✔	Identifies a localized role inside the target topology.
Local AG Server Primary	✔	✔	Verifies if the current scanned target is the primary lead host.
Availability Group Replicas	✔	✔	Full node breakdown array list mapping all active cluster replicas.
Linked Server Status	✔	✔	Flags external database target communication paths.
Command Shell Status	✔	✔	Security compliance level tracking of `xp_cmdshell` state.
SQL CLR Status	✔	✔	Verifies Common Language Runtime integration limits.
Encryption Status	✔	✔	Verifies Transparent Data Encryption (TDE) usage configurations.
Database Mirroring Status	✔	✔	Tracks database-level mirroring engine flags.
Database Replication Status	✔	✔	Confirms active structural transaction replication properties.
Maximum Server Memory	✔	✔	Memory limit allocation restrictions are measured in Megabytes.
Number of Databases	✔	✔	Active count of total databases tied to the instance loop.
Total Database Size	✔	✔	Mass file aggregation sum calculations for target cloud sizing.
AG Readable Secondary	✔	✔	Detailed permissions string for secondary read connection lines.
AG Synchronous Replication	✔	✔	Mode per replica node (Asynchronous / Synchronous).
AG Failover Mode	✔	✔	Textual description profiling the failover mode policy (Automatic / Manual).

Summary and Recommendation

Choose Agent-Based if you are planning a full-scale cloud migration that relies heavily on Application Dependency Mapping (Move Groups) and deep Microsoft SQL or licensing optimizations.
Choose Agentless if you are conducting an initial high-level TCO assessment or operating within strict security isolation zones where software installation is strictly banned.

Note: Hybrid deployments are fully supported; you can deploy the Agentless Collector for restricted boundaries while utilizing Agents for core application clusters within the same assessment portal.

If you have any queries, please get in touch with the helpdesk via our Helpdesk Portal or by email at helpdesk@cloudamize.com.