BMC health monitoring

[Vijay Khemka]

On 4/28/20, 6:23 PM, "Sui Chen" <suichen6 at gmail.com> wrote:

    Hello Vijay,
    
    I'm not aware of an BMC health monitoring package that is ready to use
    out of the box, but it appears we're thinking of similar things --
    monitor the BMC health.
    I guess there may be multiple parts to this:
    1) How does the BMC collect health metrics; this may be similar to
    profiling a regular Linux system; actually the list of metrics may be
    expanded to anything that may be obtained through Linux profilers, but
    the set of metrics that are most important might vary in different
    situations.
We can add all component of metrics to a dbus interface and frequency of 
Polling these data can be defined by config file as well as action.
    2) How to get the metric data out of the BMC to the host or some
    monitoring entity that oversees a large number of machines; depending
    on the connectivity of the BMC it may use different methods.

Rest or redfish can take these data out of BMC via bmcweb.
    
    Thanks,
    Sui
    
    On Tue, Apr 28, 2020 at 1:19 PM Vijay Khemka <vijaykhemka at fb.com> wrote:
    >
    > Is there any module which does health monitoring for BMC health. If there is none then I am proposing a phosphor-bmc-health which will monitor following component to start with and can be extended for more components.
    >
    > CPU utilization
    > Memory utilization
    >
    >
    >
    > It will take proper action like report alert or reboot system when these data will cross it’s threshold value which was configured.
    >
    >
    >
    > Regards
    >
    > -Vijay