Telco; Networking; Architecture; How to debug; Tips and Tricks
summary
Panel lays out concrete debuggability steps for containerized telco stacks: use nf_tables trace (meta nftrace) and iptables-translate to isolate rule hits and drops; enable firewalld set-log-denied for packet drop diagnostics; use OVSDPDK features (RSS queue redistribution, user‑space TSO) plus ovs prototrace, logging and coverage counters to find thread/queue hot spots; map logical to physical flows with ovn-trace and ovn‑sbctl --vflow --ovs; inspect NetworkManager with nmcli and verbose logs, apply nmstate YAML checkpoints, and capture XDP behavior with xdpdump and bpftool to spot kernel/BPF drops; example: outdated NIC firmware caused silent MTU-sized drops found via multi-point pcap.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 2:09How to debug
- 4:00TCP UDP
- 8:10NetFilter
- 12:10ODPS
- 20:49Network Manager
- 25:27EBPF and XDP
- 31:00QA
- 32:05System Type or BPF
- 34:17How to debug Telco
- 37:19Troubleshooting
- 40:29Other tools
- 41:52Networking in Telco
- 46:57Improving VPPF
- 48:07Working relationship with universities
- 49:40Questions
- 50:22Chat
- 52:27Feature dump
- 55:47Candy exchange
related talks
36mNetwork Machinery: A United-Front for Network Troubleshooting with CRDs
34mIt's Not Always the Network (But Here's How to Prove It): Kubernetes Packet...
26mPower Management for Telco Operators
29mFinding the Needle in a Haystack: Tenant-Level Network...
25mDebug Functional and Performance Issues in Rte_flow
26mNetwork monitoring with FOSS
20mCan OVS-DPDK be further optimised?
51mAdvanced approaches to traffic inspection & network diagnosis w/