Docs
test1test2
6.5
6.5
  • ElastiFlow Documentation
  • Unified Flow Collector
    • General Configuration
    • Changelog
    • Maxmind GeoIP2 and GeoLite2
    • RiskIQ PassiveTotal
    • Network Interfaces
    • User-Defined Metadata
    • Docker
    • Linux
    • Unified Flow Collector Introduction
    • System Requirements
    • Supported IEs
    • AWS VPC Flow Log IEs
    • IPFIX IEs
    • Netflow IEs
    • sFlow IEs
  • Unified SNMP Collector
    • Device Groups
    • Changelog
    • Devices
    • Downloading Definitions
    • Enumerations
    • Objects
    • Object Groups
    • User-Defined Metadata
    • Docker
    • Network Interfaces
    • United SNMP Collector Introduction
    • Linux
    • Scheduling Rediscovery
  • Monitoring ElastiFlow
    • Liveness & Readiness
    • Metrics
    • Prometheus & Grafana
  • Configuration Reference
    • YAML Configuration Files
    • Configuration Reference Overview
    • Common
      • API
      • Licensing
      • Overview
      • Logging
      • HTTP output
      • Elasticsearch output
      • Kafka output
      • Monitor output
      • OpenSearch output
      • Splunk output
      • stdout output
      • Processor
    • Unified Flow Collector
      • Overview
      • Community/Conversation IDs
      • EF_PROCESSOR_ENRICH_TOTALS_IF_NO_DELTAS
      • Overview
      • RiskIQ PassiveTotal
      • Maxmind
      • User-Defined Metadata
      • Overview
      • Overview
      • User-Defined Metadata
      • Overview
      • Benchmark Input
      • Netflow/IPFIX/sFlow (UDP)
      • Licensing
      • Decoder/Processor
      • Sample Rate
      • Configuration Changes
    • Unified SNMP Collector
      • User-Defined Metadata
      • Overview
      • Licensing
      • SNMP Poller
      • EF_PROCESSOR_SNMP_ENUM_DEFINITIONS_DIRECTORY_PATH
  • API Reference
    • API Reference Overview
    • SNMP Operations
  • Data Platforms
    • Elastic
      • Basic Cluster
      • Advanced Cluster
      • Single Server
      • Multi-Tier Cluster
      • Single "Lab" Server
      • Elasticsearch
      • ElastiFlow vs. Filebeat and Logstash
      • RHEL/CentOS
      • Ubuntu/Debian
      • Kibana
      • ML
        • Network Security
        • Machine Learning
        • Availability
          • Network Availability
          • DHCP
          • LDAP
          • DNS
          • NTP
          • RADIUS
          • TCP Sessions
        • Network Security Activity
          • Rare Autonomous System
          • Network Activity
          • Rare Conversation
          • Rare Geolocation
        • Network Security Brute Force
          • Brute Force CLI Access
          • Brute Force Remote Desktop Access
          • Brute Force Attacks
        • Network Security DDoS
          • Denial-of-Service
          • ICMP Flood Attack
          • SYN Flood Attack
          • TCP DDoS Attack
          • UDP Amplification Attack
        • Network Security Recon
          • ICMP Scan
          • Reconnaissance
          • Port Scan
        • Performance
          • Unusual ASN Traffic Volume
          • Unusual Network Interface Traffic Volume
          • Network Performance
    • Opensearch
      • Dashboards
      • Auth Sig V4
    • Splunk
      • Default Search Macro
      • Configuring Data Input & Index
      • Splunk App Installation
    • Output Configuration
  • Additional Guides
    • Catalyst (sFlow)
    • FortiGate
    • hsflowd
    • Configuring Flow Sampling on Juniper Routers
    • Junos OS (sFlow)
    • MikroTik RouterOS
    • OpenWRT (softflowd)
    • Ubiquiti EdgeRouter
    • SonicWall
    • Junos OS
    • Extending SNMP Device Support
    • Flow Device Support Overview
    • SNMP Device Support Overview
    • Generating A Support Bundle
  • FAQ
    • Flows stopped showing up in Kibana (Disk(s) Full)
    • Common reasons why you have discrepancies between ElastiFlow data & reality
    • What Are Snapshots?
    • Importing the wrong dashboards (No data)
  • Knowledge Base
    • Config
      • Elasticsearch Authentication Failure
      • CA Certificate Path Incorrect
      • license/error Invalid Segments
    • Flow
      • Bidirectional Flow Support
      • Configure the UDP Input
      • Flow Records Not Received
      • Netflow v9/IPFIX Template Not Receieved
      • Unsupported sFlow Structures
    • General
      • License Has Expired
      • License Agreement Not Accepted
    • Install
      • .deb Upgrade Fails File Overwrite
    • Operation
      • Flow Collector Queues 90% Full
      • Dashboard Updates
      • Change elastiflow-* Index Name?
  • Elastic Stack Deployment
  • Download Links
Powered by GitBook
On this page
  1. Data Platforms
  2. Elastic
  3. ML
  4. Availability

TCP Sessions

The Transmission Control Protocol (TCP) is one of the core protocols of the Internet Protocol Suite. It is designed to provide a reliable, ordered, and error-checked delivery of a stream of data between applications running on hosts communicating over an IP network. TCP is fundamental to modern internet-based communication, supporting a wide range of applications from web browsing and email to file transfers and streaming services. Its primary purpose is to ensure that data packets sent from one end of the network are received accurately and in the same order by the intended recipient at the other end. This is achieved through various mechanisms like packet segmentation, acknowledgement of received packets, and retransmission of lost packets, making TCP a cornerstone protocol for reliable communication over the internet.

TCP operates as a session-oriented protocol, meaning it establishes a connection between two endpoints (a client and a server) before data transfer begins, and maintains this connection until the communication session ends. The protocol uses various flags (such as SYN for initiating connections, ACK for acknowledging received packets, and FIN for closing connections), options (like window scaling and selective acknowledgments), and other information within the TCP header to manage and control the flow of data. By analyzing these components in the TCP headers of packets traversing a network, network engineers and IT professionals can detect disruptions in TCP services. For instance, repeated retransmission requests or a high number of reset (RST) flags might indicate connection issues or network congestion. Similarly, unusual patterns in TCP flags or options can signal potential security threats like TCP hijacking or Denial of Service (DoS) attacks. Therefore, monitoring and analyzing TCP traffic is crucial for maintaining the stability, performance, and security of networked communication systems.

Failed TCP Sessions

The "Failed TCP Sessions" anomaly detection job is designed to monitor and identify instances of an unusually high volume of failed TCP (Transmission Control Protocol) sessions within a network. TCP sessions are considered failed when they are unable to successfully establish a connection. In typical network operations, the number of failed TCP sessions is relatively low compared to successful ones, as TCP is designed to be a reliable, connection-oriented protocol.

An unusually high volume of failed TCP sessions can indicate several potential issues in the network:

  • Network Congestion or Unreliable Connections: One of the primary reasons for an increase in failed TCP sessions could be network congestion or unstable connections. Heavy network traffic or poor quality of service can lead to packet loss, high latency, and jitter, which in turn can disrupt the normal TCP three-way handshake process or cause premature session terminations.

  • Server Overload or Failures: If servers that are the endpoints of these TCP sessions are overloaded with requests or experiencing failures (either hardware or software), they may be unable to respond to connection requests or maintain existing connections, leading to an increase in failed sessions.

  • Security Incidents: A high rate of failed TCP sessions could also be indicative of security threats, such as Denial of Service (DoS) or Distributed Denial of Service (DDoS) attacks. These attacks typically involve overwhelming a target with superfluous requests, aiming to disrupt normal traffic and cause legitimate connection attempts to fail.

  • Network Configuration or Hardware Issues: Misconfigurations in network settings, such as incorrect firewall rules or routing problems, can block or misroute TCP traffic, leading to session failures. Similarly, malfunctioning network hardware like routers, switches, or cables can also contribute to this problem.

Attributes

Attribute
Information

Analysis

temporal

Downloads

Schema
Perspective
Link

CODEX

edge

CODEX

inbound

CODEX

outbound

CODEX

private

ECS

edge

ECS

inbound

ECS

outbound

ECS

private

PreviousRADIUSNextNetwork Security Activity

elastiflow_codex_avail_tcp_sess_fails_edge
elastiflow_codex_avail_tcp_sess_fails_in
elastiflow_codex_avail_tcp_sess_fails_out
elastiflow_codex_avail_tcp_sess_fails_priv
elastiflow_ecs_avail_tcp_sess_fails_edge
elastiflow_ecs_avail_tcp_sess_fails_in
elastiflow_ecs_avail_tcp_sess_fails_out
elastiflow_ecs_avail_tcp_sess_fails_priv