Aviz Service Node details DPI, decapsulation, deduplication, Kafka KPIs
Aviz Service Node (ASN) provides IP traffic observability by tracking millions of sessions, performing application identification and tunnel decapsulation, and exporting per-session and per-application KPIs to Kafka at 5-second intervals, supporting troubleshooting and analytics in data center, edge, enterprise, and FTTH networks.
Research Overview
The blog describes Aviz Service Node (ASN) as a high-performance observability platform for data centers, edge and distributed sites, enterprise and campus networks, and FTTH environments.
It focuses on extracting per-session metadata, deriving time-bound KPIs, and using streaming outputs to support performance monitoring, capacity planning, and day-2 operations.
Key Findings
ASN analyzes high-volume packet streams by tracking millions of IP sessions in memory using IP tuples to create indexed metadata per session for high-throughput processing.
The DPI engine identifies 500+ applications in the current scope, with a roadmap target of 1000+ applications, using multiple classification techniques.
Technical Breakdown
For protocol inspection, ASN performs selective header stripping and decapsulation to reveal inner payloads for DPI and root-cause analysis.
The blog lists support for GRE, ERSPAN, VXLAN, PPPoE, L2TP, and IPoE encapsulations.
KPI Calculation, Streaming, and Alerting
ASN exports enriched APP, KPI, and session records to external platforms via Kafka, with KPI aggregation and streaming occurring every 5 seconds for real-time observability.
Timestamping supports time-window analytics, and priority events such as session deletion, release, or handover trigger immediate Kafka notifications for rapid triage.
Packet Deduplication and Live Capture
The blog describes configurable packet deduplication to reduce duplicate packets that can inflate bandwidth and skew metrics, especially in large transfers and mirrored streams.
Controls include selecting tap/SPAN/ERSPAN/VXLAN sources for the dedup scope, choosing an anchor and offset, and setting a time/byte window size for line-rate efficiency.
For troubleshooting in high-throughput data center environments, ASN supports live packet capture alongside line-rate processing for debugging, validation, and ad-hoc analysis.
Blog Signals brief is a fact-based summary of the vendor blog.