Self-Health Monitoring

From 10.3.2, users can benefit from the newly introduced 'Self-Health Monitoring' or the 'Self-Healing' feature that monitors the health of SpectroSERVER as shown in the screenshot below. This feature ensures SpectroSERVER runs smoothly in adverse conditions such as faulty devices, interface issues, CPU, and memory hassles.
Self Healing.JPG
Following are the capabilities supported as part of this release:
  1. Detect CPU's over capacity
    - SpectroSERVER (SS) creates perf dmps for continuous high CPU utilization, for a designated time period. The threshold values are configurable. The dmp files are helpful in case the SS runs into performance issues.
  2. Detect memory over capacity
    - Kernel crashes SS if the system runs out of available or swap memory. This issue is now addressed and the SS shuts down gracefully, in case the available and swap memory is less than 100 Meg (configurable).
  3. Excessive interface re-configurations handling
    - Previously SS raised ‘Excessive Interface Reconfigure' alarm on a device if six interface re-configurations occur within 31 minutes. These frequent interface re-configurations adversely affect the SpectroSERVER's performance. To address this issue, from 10.3.2 onwards, if this option is enabled, SS disables auto-reconfiguration on such devices and waits for an age-out time (configurable to 1 hour by default). After the age-out time expires, SS enables the reconfiguration.
  4. High port count handling
    - If this option is enabled, SS restricts the continuous re-configurations on high port count devices by scheduling it for every 15 minutes (configurable). By default, devices with port count of more than 1000 are identified as high port devices and this value is configurable.
It is recommended that you enable all the capabilities/options that are mentioned above on all the SpectroSERVERS.
Note that from 10.4.2 onward this functionality is enabled by default.