Real BMC emulators. Real IPMI commands. Real data center skills.
Manage servers remotely the way data center engineers do. SSH into BMC consoles, run IPMI commands, read sensors, control power, pull event logs, and automate with Redfish REST APIs. Each lab connects you to a live BMC emulator alongside your Firecracker workstation.
10 published lessons across 10 chapters, about 8 hours of work. Every lab runs on a real virtual machine in your browser, and checks your work on the machine itself.
Chapter 1 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller beside it: reaching a BMC over the network with ipmitool, mc info and the GUID, chassis and power state, lan print and the cipher table, the factory user table and the fleet standard (password rotation, an operator account, the anonymous slot), the SEL, the SDR and sensor thresholds, FRU, SOL and the watchdog, the two access paths, and the handover.
Chapter 2 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller that owns a real QEMU guest beside it: the power state machine over the network with ipmitool, chassis power status, on and off, the escalation from a graceful request to reset to cycle, watching a node boot on the console over SOL, the power restore policy, and the vendor power verbs, restart cause and diagnostic interrupt a real Dell BMC adds.
Chapter 3 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller beside it: reading the System Event Log with ipmitool, sel info, sel list and sel elist, slicing an eight-event overnight incident by subsystem, naming an uncorrectable-memory-ECC root cause and separating it from a survivable power-supply red herring, the Pre-Init clock and why a real BMC syncs it with NTP, one record in full, the genuine Dell iDRAC event-log reference and the vendor dialects, and closing the task without clearing the evidence.
Chapter 4 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller beside it, polling five live sensors once a second: reading them three ways with sensor list, sdr list and sdr elist, one sensor in full with sensor get, the six threshold columns and what na means, filtering with sdr type, a sensor baseline, a fan failing at 600 RPM against a healthy 5400, an exhaust temperature that really crosses its critical line while the lab runs, the threshold events the controller raises by itself, the fan to heat correlation, the genuine Dell R730 sensor reference, and the verdict on whether the node comes down.
Chapter 5 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller carrying a real multi-device FRU tree: walking the tree with fru print and reading each device on its own, extracting a serial the way an RMA form wants it, looping the device ids, the FRU Device Locator records that make discovery work at all, the system GUID and its SMBIOS encoding warning, the entity map that ties a sensor to a part, a sensor presence check, and the two identification commands this controller does not implement, captured as errors and told apart by their completion codes, with the genuine Dell R730 reference beside them.
Chapter 6 on a real Ubuntu 24.04 workstation called mgmt with a real IPMI 2.0 controller whose accounts are real: the user table and what every column means, creating an account in four commands and proving it establishes an RMCP+ session, the refusal a wrong password returns, retiring a vendor account left over from a contract that ended and proving the door is shut, why those two refusals look identical and how to tell them apart, the channel and its cipher suites, the management network, what a VLAN tag costs on real hardware, and the three writes this controller refuses, with the genuine Dell R730 user table and network beside them.
Chapter 7 on a real Ubuntu 24.04 workstation with a real IPMI 2.0 controller owning a real virtual machine: what a serial console carries and why it survives the loss of the host network, ruling out a hardware fault before touching the console, the single-session payload and how to reclaim one that is held, capturing a whole boot to a file with script(1) and converting it to readable text, reading the firmware handoff, the kernel banner and the userspace phase out of a capture, the one line that explains a host with no network, the console parameters this firmware does not implement, the vendor escape sequences, and a written diagnosis supported by the capture.
Chapter 8 on a real Ubuntu 24.04 workstation with a real IPMI 2.0 controller: boot options are eight numbered parameters rather than one setting, and this firmware implements two of them. Read the ones that answer, capture the refusals from the ones that do not, learn what chassis bootdev writes from its own error, prove at the raw command level that the gap is the firmware, watch the machine choose a boot device on its own console, read a genuine Dell boot flags decode field by field, and write the procedure and the report a change request needs when the equipment cannot do what was asked.
Chapter 9 on a real Ubuntu 24.04 workstation with a real IPMI 2.0 controller: a watchdog is a countdown the host has to reload, and the controller acts on the silence when the reloads stop. Read a timer that was never set, collect the refusal that gets, name the five stages and the four timeout actions, arm one with a raw command because ipmitool cannot, prove that setting a timer does not start it, watch a real expiry latch its flag and write its own record, let one power the host down, clear the latch, and write the documentation an operator needs before arming a watchdog in production.
Chapter 10 on a real Ubuntu 24.04 workstation with a real IPMI 2.0 controller and a managed host that can answer from inside itself: the same controller reached over the network with an address and a sign-in, and from the host it manages with neither. Read the controller both ways and compare four identity fields to establish it is one device, watch the three IPMI kernel modules build the path and create /dev/ipmi0, trace how firmware described the KCS interface twice and the driver resolved it once, read the LAN configuration from both sides and find the gateway and VLAN parameters the in-band reading stops short of, read the user table from the host with no sign-in at all, and record which fault leaves which path.
BMC & Server Management has 10 published lessons, roughly 8 hours of guided work, spread across 10 chapters.
No. Every lab boots a real virtual machine in your browser, with a real Linux shell. There is nothing to install, and nothing on your own machine to break.
Labs in this track are rated beginner, intermediate. The track starts from the beginning, so no prior experience is assumed.