jda/bongo

Gather metrics from Canopy CPE and present them in a useful manner

Bongo: Recording the beat of your Canopy jungle

General idea

Given a list of AP LUID Subnets: Discover hosts within subnet that are alive Poll every host for the vital statistics Assume every radio has default community Record data in RRD - one per MAC address Subfolders by MAC prefix

Assumptions

All customer SMs on a AP have the same SNMP community Your SM Backhauls should not, instead they should be graphed in your EMS/Cacti/Etc, not Bongo. Disk is cheap. Allocate RRD to store stats at 5-minute intervals for 5 years.

ToDo

Record current SM stats in per-SM sqlite DB Record current SM stats in per-run sqlite DB for on the fly queries

Split out config logging into seperate program that runs less often.

2-Stage polling - Poll for SW version. On response Poll OIDs relevent on version. Also means frwer outbound packets.

Config file

MIBs Tracked

• RSSI: 1.3.6.1.4.1.161.19.3.2.2.2.0
• Jitter: 1.3.6.1.4.1.161.19.3.2.2.3.0
• dBm: 1.3.6.1.4.1.161.19.3.2.2.21.0
• Parent AP: 1.3.6.1.4.1.161.19.3.2.2.9.0
• NAT mode: 1.3.6.1.4.1.161.19.3.2.1.19.0
• RFC1213-MIB::ifPhysAddress.2
• RFC1213-MIB::ifInOctets.2
• RFC1213-MIB::ifOutOctets.2
• RFC1213-MIB::ifInUcastPkts.2
• RFC1213-MIB::ifOutUcastPkts.2
• RFC1213-MIB::ifInNUcastPkts.2
• RFC1213-MIB::ifOutNUcastPkts.2
• RFC1213-MIB::ifInErrors.2
• RFC1213-MIB::ifOutErrors.2
• RFC1213-MIB::ifInDiscards.2
• RFC1213-MIB::ifOutDiscards.2