XDR Analyst flashcards
XDR Analyst · 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.
- XDR definition
- A platform that collects and correlates telemetry from endpoint, network, identity, email, and cloud sources into unified detections and response actions, rather than operating as a single isolated sensor.
- Normalization schema
- A common field model such as ECS or OCSF that maps vendor-specific log fields into consistent names, enabling detections and correlations to work across heterogeneous data sources.
- Entity resolution
- The process of mapping raw identifiers like hostnames, IPs, and usernames to canonical assets and identities, preventing one intrusion from fragmenting into many disconnected alerts.
- Telemetry health metric
- A coverage measure tracking the percentage of enrolled endpoints reporting recently and flagging ingestion gaps or parser failures, since missing data silently blinds detections.
- Process lineage
- The parent-child relationship chain of process execution, a core EDR detection primitive that reveals malicious execution flows single-event rules miss, such as Office spawning PowerShell.
- LOLBins
- Legitimate signed operating system binaries such as certutil, mshta, rundll32, and regsvr32 that attackers abuse for download, execution, or persistence; detection relies on command-line arguments rather than binary reputation.
- LSASS credential access
- Unauthorized processes opening handles to lsass.exe with read rights, or tools like comsvcs.dll MiniDump being invoked, indicating credential dumping that precedes lateral movement.
- Endpoint isolation
- A reversible containment action that blocks most network traffic to and from a host while preserving the agent's management connection, preferred over shutdown because it retains volatile evidence.
- Sysmon Event ID 1
- Records process creation with image path, command line, parent process, user, and hashes; the highest-value Windows endpoint telemetry source for behavioral detection.
- Beaconing detection
- Identifying command-and-control by measuring inter-arrival time variance of repeated connections to the same destination; low variance with consistent payload sizes over many connections is suspicious.
- DNS tunneling indicators
- High volumes of TXT or NULL queries, unusually long encoded subdomain labels with high entropy, and many unique subdomains under one authoritative domain.
- Cloud control-plane logs
- Audit records of management API calls such as CloudTrail or Azure Activity Log; high-value events include CreateAccessKey, AttachUserPolicy, and StopLogging.
- Flow log limitations
- VPC and NSG flow logs provide 5-tuple records with byte counts but no payload, and sampling means absence of a flow is not proof of no traffic.
- East-west traffic
- Internal lateral communication between hosts; anomalies on SMB, RDP, WinRM, and Kerberos between workstations often indicate post-compromise movement rather than normal operations.
- Impossible travel
- A detection comparing consecutive successful authentications to compute implied velocity exceeding physical travel; false positives arise from VPN egress, carrier NAT, and cloud proxies.
- MFA fatigue attack
- Repeated push notification prompts sent to a victim until they approve; detected by joining many denied MFA events to a later success from the same user and source.
- DMARC alignment
- A policy that requires SPF or DKIM to align with the visible From domain and instructs receivers to none, quarantine, or reject failing mail.
- Business email compromise
- Fraud using lookalike or new domains, reply-to mismatches, urgency language, and thread hijacking to trick recipients into payments or credential disclosure.
- Token theft
- Stealing OAuth refresh tokens to bypass MFA; indicators include anomalous token use, new device registrations, and mailbox rule creation after a suspicious sign-in.
- Intelligence pyramid
- A model ranking indicators by durability: hashes and IPs are cheap and short-lived, while tactics, techniques, and procedures are expensive for adversaries to change.
- Indicator TTL
- An expiry assigned to threat intelligence indicators, requiring revalidation before use because stale IP and domain indicators cause false positives as infrastructure is repurposed.
- MITRE ATT&CK
- A knowledge base of adversary tactics and techniques used to map detections to behaviors, enabling coverage gap analysis across the attack lifecycle.
- Sigma rules
- A vendor-neutral detection rule format that compiles to multiple SIEM query languages, supporting portable detection-as-code with version control and peer review.
- Stack counting
- A hunting query pattern that groups events by a field combination and sorts by count, surfacing rare parent-child process pairs or uncommon command lines.
- Purple teaming
- Executing known adversary techniques with tools like Atomic Red Team or CALDERA and measuring which detections fire, producing a concrete list of coverage gaps.
- Severity calculation
- Alert severity derives from asset criticality, identity privilege, detection confidence, and blast radius, not from the alert source alone.
- Containment ordering
- Actions ordered by reversibility: revoke sessions and reset passwords, isolate the endpoint, disable the account, and reimage the host only as a last resort.
- Volatile evidence
- Data such as memory contents, running processes, and active network connections that is lost on shutdown and must be captured before containment or reboot.
- SOAR approval gate
- A conditional step requiring human authorization before destructive actions like deletion, reimage, or mass changes, while reversible actions execute automatically.
- MTTD and MTTR
- Mean time to detect and mean time to respond; tracked per detection rule to identify noisy, slow, or ineffective detections for tuning.
- Blast radius
- The scope of systems, identities, and data an intrusion can reach from a compromised entity, a key input to incident severity and containment urgency.
- Mailbox persistence
- Attacker-created forwarding rules, delegate access grants, and OAuth consent grants that survive password resets and must be enumerated during containment.