monitor all your stuff RIGHT NOW!!

About this video

Chuck deploys Progress WhatsUp Gold — the first monitoring tool he used on the help desk, now offered in a free edition — to stop guessing what his servers, routers, switches, storage and GPU-heavy AI server are doing. He explains monitoring in layers: ping (ICMP echo request/reply) only proves a device is awake, while SNMP and Windows-only WMI reveal whether it is actually working — CPU, memory, disk space, temperature and interface traffic, queried via numeric OIDs, with vendor-specific MIB packs adding proprietary metrics. He configures SNMP on Linux, warns that v1/v2c community strings (default ‘public’) travel in plain text and that read-write access can even change a hostname, and recommends SNMPv3 with encryption for production. He also preps Windows via firewall rules for ICMP/WMI and an admin account, installs WhatsUp Gold, runs discovery with SNMP/WMI/SSH credentials, and then shows dashboards and maps, a custom HTTP monitor for a service on port 3000, a Slack alert policy for down/up states, and custom SSH performance monitors that poll GPU temperatures — arguing monitoring is a core, resume-worthy IT skill.

Key takeaways

• Ping only answers ‘is it up?’; SNMP and WMI answer ‘is it working?’ by exposing CPU, memory, disk, temperature and interface data via OIDs.
• SNMP polls by default roughly every 10 minutes on UDP port 161; SNMP traps let a device push an alert immediately when a threshold is crossed instead of waiting to be polled.
• Treat community strings as passwords: the default ‘public’, plain-text v1/v2c and read-write strings are dangerous — SNMPv3 adds encryption, username/password authentication and restricted views.
• The free edition monitors up to 10 devices without per-sensor limits; Chuck cites install needs of a quad-core processor, at least 8 GB of RAM and at least 25 GB of storage.
• Custom monitors extend coverage to almost anything: an HTTP check on port 3000 for Open WebUI and SSH commands polling GPU temperature, tied to action policies that alert via Slack and other actions.

Episode details: published May 16, 2024 · 32:52 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are NetworkChuck’s own claims, shared for education — they are not typical or guaranteed.

← Previous: host ALL your AI locally · Next: I use this everyday →

All NetworkChuck videos · NetworkChuck profile · NetworkChuck on YouTube