Category:INFOSYS DGRAM ELK

Training on INFOSYS DGRAM ELK KT Part 1. Dated 05 June 2026 - Tetra Support Staff - Prashant[edit]

Video summary: Part 1 of the INFOSYS DGRAM ELK KT (see Part 2, 6IrkurBVESg, for the deeper data-flow explanation). Gives the fuller device-coverage picture for the 19-node observability cluster: roughly 400-900 monitored assets spanning Linux/Windows servers, network devices (Juniper routers, Aruba switches), storage arrays (NetApp done via SNMP; a Pure Storage array not yet integrated because the vendor declined to cooperate), and a range of security devices -- Fortinet sandbox, web proxy, DLP, and F5 load balancers, only partially onboarded because the client's security team hasn't enabled SNMP access on some of them yet. Explains the two integration patterns used: for devices Elastic has an official "out-of-the-box" integration for (e.g. Palo Alto next-gen firewalls), logs flow in with minimal setup once Elastic Agent is pointed at them; for devices that can't be integrated directly (older network/security gear), an intermediate collector server receives the device's SNMP/syslog output and runs Elastic Agent itself to forward that on into the cluster.

Training on INFOSYS DGRAM ELK KT Part 2. Dated 05 June 2026 - Tetra Support Staff - Prashant[edit]

Video summary: Part 2 of a KT session on the ELK-based observability platform built for the Infosys DGRAM project. Explains the basic data flow taught by analogy to Prometheus+Grafana: an Elastic Agent installed on each monitored server collects logs/metrics and ships them to Elasticsearch, which Kibana then queries to render dashboards (Logstash sits optionally in between as a filtering/transform pipeline, but isn't heavily used here since Elastic Agent already does most of that collection work natively). Describes the two clusters built for this client: a smaller "search cluster" (8 nodes: 3 combined master+data Elasticsearch nodes, 1 machine-learning-role node, 2 Kibana, 2 Logstash) and a much larger 19-node "observability cluster" (11 Elasticsearch nodes including warm-tier nodes plus 3 machine-learning-role nodes, 2 Kibana, 2 Fleet servers, 2 APM nodes, 2 Logstash). Also introduces Fleet, Elastic's centralized agent-management service, which lets them configure and monitor every Elastic Agent across the fleet from one place (a handful of commands to enroll a new agent) rather than configuring each server's agent by hand.