Editing Training 2022 Linux team/ (section)

==Training session Tetra team on Infosys ITD Nagios understanding. Dated 26 Nov 2022 - Tetra Support Staff - Biswajit Banerjee==

{{#ev:youtube|SB7wZlQc4GU|640}}

'''Video summary:''' "Training session Tetra team on Infosys ITD Nagios understanding" -- a detailed handover of the ITD monitoring platform built on NagiosXI, covering roughly 1200 devices. Demonstrates the live monitoring web dashboard/console and reviews original architecture documentation. Explains SNMP as the core monitoring protocol -- pulling metrics like traffic, memory, and port status from network devices and storage without needing local agents -- and that "NRPE proxying" (routing monitoring traffic through an intermediary) is only needed for devices not directly network-reachable or on a segmented network, which wasn't required here since ITD's devices were all directly reachable. Covers WMI (Windows Management Interface) as the equivalent mechanism used for Windows hosts. Walks through the range of what's actually monitored: network switches/devices (fan speed, dual power-supply failover status, temperature, per-interface bit/error rates); physical server health via IPMI/BMC-style out-of-band management ports (remote OS install, hardware status, disk/memory failures) on IBM AIX and other platforms; virtualization platforms (VM failure detection, storage consumption, snapshot/backup counts); and Hitachi enterprise storage (capacity, per-volume utilization, disk failure) queried via a dedicated query package rather than plain SNMP. Notes the full onboarding of this many unique device types took roughly 5-6 months of iterative work, and that AIX support specifically required NRPE-based agents rather than SNMP given the platform's specifics (RPM package copied and installed per-client machine, with documented troubleshooting steps for common install errors). Explains the team's standard troubleshooting handoff pattern for network-device monitoring gaps: check the target device output first using a documented reference command/expected-output pair; if no output comes back, escalate to the relevant network team (separate teams typically own switches/routers, firewalls, and other device classes respectively) since a missing SNMP response at that layer isn't something the monitoring team can fix themselves. Closes noting that despite the platform's scale, day-to-day support tickets are rare because the client's own on-site team (named contacts including Sudhanshu and Ashish) handles ground-level operations, with Tetra mainly stepping in for upgrades or platform-level issues.