banshee is under active development: collectors will keep changing, and a version you publish will likely need adjustments. For stable use, wait for version 1.0.0.

Hardware

banshee/hardware-collection/hardware/ibm/ssh/hmc/

serviceable-events

The open serviceable events of an IBM HMC, the hardware problems of its servers waiting for service, by SSH.

  • hardware
  • ibm
  • power
  • hmc
  • serviceable-events
  • lssvcevents
  • ssh

by banshee0 installations1 version

v1Latest

Published on October 27, 2026. First version.

Open serviceable events of an IBM HMC, by SSH

Returns the number of open serviceable events of a Hardware Management Console (HMC), and their problem numbers: the hardware problems the IBM Power servers it manages have reported, waiting for service. One watch is offered. It says why managed-systems sees an attention LED lit.

Versions

Written for the command lssvcevents of the HMC as IBM documents it for the HMC of Power10 (HMC V10), with the option -F and the filter status; whether earlier HMCs take them is not checked. Hence the common level, no other version being known to differ. The HMC REST API serves no serviceable event: hence SSH, under the folder ssh/hmc beside the REST collectors.

How it works

The collector runs, on the HMC's command line, by SSH, IBM's example for the open serviceable events:

lssvcevents -t hardware --filter "status=open" -F problem_num:status

It only lists: each open event on a line, its problem number and its status separated by :, 1234:open.

  • open is the number of these lines; problems their problem numbers, joined by , , '' when there is none.
  • A line of any other form stops the collector misconfigured, with the line: an output it does not understand never reads as "no event".
  • A command that exits with another code than 0 stops it misconfigured, with what the command wrote: the account may not run it.

The command is the one IBM documents, written whole in the collector; it changes nothing on the HMC.

Prerequisites

  • The HMC's address as the host's address, reachable from the poller by SSH (22), its host key pinned in protocol:ssh:fingerprint or trusted on first connection in the tab "Security", as for any SSH probe.
  • An HMC account in protocol:ssh:user with protocol:ssh:password or protocol:ssh:private-key, allowed to run lssvcevents -t hardware. The predefined hmcviewer role may not: IBM's table of the HMC tasks and roles gives the Serviceable Events Manager (lssvcevents, chsvcevent) to the Super Administrator (hmcsuperadmin) and the Service Representative (hmcservicerep) only, both able to change the HMC. The narrowest account is one of a customized task role, made on the HMC from hmcviewer with the Serviceable Events Manager added; that task also allows chsvcevent, which closes or approves an event, and no narrower task exists.
  • Remote command execution enabled on the HMC (the HMC task "Enable Remote Command Execution").

Outputs

NameLabelUnitPer instance
openOpen serviceable events—no
problemsProblem numbers—no

Watches

NameTitleKindDefaultsChecked by default
open-eventsServiceable events openhigh threshold, on openwarning above 0yes
  • An event stays open until it is closed on the HMC, after service; open-events stays raised until then.

Limits

  • What lssvcevents writes when no event is open, and its exit code then, are not in IBM's documentation. If it writes a message rather than nothing, or exits with another code than 0, the probe stops misconfigured on that message instead of returning 0: never a green it has not read. The first run on a healthy HMC says which.
  • The reference code, the system and the text of an event are not read: IBM documents the names of the attributes problem_num and status only.
  • Not checked against an HMC: the command, its options and its output are those of IBM's manual page of lssvcevents (HMC V10); the roles are those of IBM's table "HMC tasks, user roles, IDs, and associated commands".