The TDI Alerts Server has the ability to detect certain system conditions and to execute user procedures when these conditions occur.
The alert conditions are selected by a set of "alert_*" files which specify the conditions and thresholds for an alert. Alerts may be scheduled by day of the week and time of day. When an alert condition is detected, the TDI Toolkit calls a command macro, "manage_*.cm", where "*" is the object being watched, e.g. module, disk, device, etc. TDI provides examples for each macro. However the user is responsible for the macros and their contents.
The macros can be easily configured to work with the Syllog, Inc. SFAX system for numeric or alphanumeric paging. They may also be configured with a package for SNMP for network-based problem management.
The conditions that may generate alerts are:
Module:
Empty idle is below a user-specified value.
Page faults per second is above a user-specified value.
Interrupts per second is above a user-specified value.
Paging Space available is below a user-specified value
Disk: Disk is simplexed.
File partition percent available is below a user-specified value.
Paging partition percent available is below a user-specified value.
Depletion - file partition percent available is estimated to go to zero within a user-specified time.
File:
File does/does not exist.
File size outside of range.
File growth outside of range.
File lockers outside of range.
Process:
Process does/does not exist.
Process busy, CPU busy, memory, or paging rate outside of range.
Batch:
Queue missing or not started.
Job is/is not in queue.
Job is/is not executing.
Job is in queue too long.
Job is a duplicate.
Job has an invalid output path.
Queue:
Number of messages in the queue is above a user-specified value.
Oldest message in the queue is older than a user-specified value.
Device:
DSR, Carrier or controlling process is down.
Throughput or error rate is outside of range.
Log Contents:
Log or file record contains a user-specified string.
Temperature:
Computer room temperature too warm.
Electrical:
Power failure event.
|
