SCTP Load Balancer Session Consistency via Intelligent Hashing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing load balancing techniques in visibility networks, which rely on 5-tuple hashing, are inadequate for SCTP traffic due to multi-homing capabilities that cause packets to be rerouted to different probes after network failures, leading to undesirable destination changes and potential probe overload.
Innovation Solution
Implementing an SCTP load balancer that receives replicated SCTP packets, recovers SCTP messages, and maps them to egress ports based on decoded parameters and user-defined rules, ensuring all traffic for a single communication session goes to the same probe, even after failovers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If 5-tuple hashing is used for load balancing, then traffic distribution is simple and fast, but SCTP multi-homing cause packets to be routed to different probes after failover, breaking session consistency
Solution Approach 1:
The patent changes the hashing parameters from standard 5-tuple to include SCTP-specific parameters (association ID, stream ID, message sequence number) to maintain session consistency while still enabling load balancing. This allows the system to differentiate between multiple SCTP associations and ensure all packets for a given association are routed to the same probe even after failover events.
Solution Approach 2:
The patent introduces an intermediary load balancer component that sits between the network core and analysis probes. This intermediary performs intelligent SCTP-aware hashing and session tracking, mediating between the incoming traffic and the probe array to ensure session consistency is maintained while still distributing load across multiple probes.
2Reliability
If all traffic between two endpoints is routed to a single probe, then session state analysis is simplified, but the probe may become overloaded when traffic volume is particularly large
Solution Approach 1:
The patent segments the traffic flow by applying hierarchical hashing: first by endpoint pair (to maintain session grouping), then by SCTP association ID (to distribute different associations within the same endpoint pair to different probes). This segmentation allows session state analysis to be maintained within each probe while distributing the overall load across multiple probes, preventing any single probe from becoming overloaded.
3Productivity
If load balancing distributes traffic across multiple probes, then no single probe becomes overloaded, but session state information may be lost when packets are hashed to different probes
Solution Approach 1:
The patent implements preliminary session tracking and association mapping before traffic is distributed to probes. The load balancer pre-establishes which SCTP associations are mapped to which probes based on hashing rules, and maintains this mapping information. This preliminary action ensures that even as traffic is distributed across multiple probes, session state information is preserved because the system knows which probe is responsible for which association in advance.
Data Source
AI summary
Techniques for enabling offline, intelligent load balancing of Stream Control Transmission Protocol (SCTP) traffic are provided. According to one embodiment, a load balancer can receive one or more SCTP packets that have been replicated from a network being monitored. The load balancer can further recover an SCTP message from the one or more SCTP packets and can map the SCTP message to an egress port based on one or more parameters decoded from the SCTP message and one or more rules. The load balancer can then transmit the SCTP message out of the egress port towards an analytic probe or tool for analysis.