Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

This is a brief overview of the major changes between opEvents releases.

opEvents 2.4.2

This major release of opEvents introduces several new features and bugfixes, published 1 November 2018

(warning) opEvents 2.4.2 minimum MongoDB version is 3.2

  • New in-app editor for Event Actions and policies, this allows quick editing of you EventActions.nmis config, this can be found under "System" > "Edit Event Actions"
    • Syntax Highlighting for Event Actions
    • Validate your changes at any stage
    • Console output for event actions validation
    • Embedded documentation
    • Each save generates a new backup of your last revision these are saved in <omk_dir>/backups/EventActions.nmis.*
  • Node context links are now viewable in opEvents. Docs for how they are used inside nodes at NMIS Linking Nodes to other systems
    • Node refresh from NMIS will be required to import these properties if changed on the node.
    • These show in the navigation bar for:
      • Node & Event Context
    • These show as links inside the data tables for:
      • Dashboard Screen
      • Events List & Current Events
  • More robust handling of system overload situations
    • The main dashboard screen now warns if the event or action queues
      grow too large.
    • Any processing of events that have been stuck in the incoming queue beyond
      opevents_max_event_age (default: 2 hours) is now aborted and an appropriate
      entry is written to the raw log.
    • A new diagnostic API endpoint at omk/opEvents/health.json was added,
      which reports the opEvents daemon states and some metrics on queued overdue
      events and actions.
  •  Resource safety limits for request periods
    • opEvents now by default sets the new configuration option
      opevents_max_period_value to 30 days, which ensures that no periods greater
      than that are selectable from the GUI. otherwise, selecting overlong request
      periods may cause memory exhaustion for both MongoDB daemon as well as the
      opmantek webserver workers.
  • opEvents now installs a purging policy for old data by default
    (unless the admin answers 'no' when the installer prompts)
  • The events dashboard pages now select a default display interval
    that extends slightly into the future (instead of cutting the interval
    off before the current time)
  • The Opmantek web server now supports resource limits to alleviate out-of-memory situations
    • The two new configuration options omkd_max_requests and omkd_max_memory
      can be used to ensure that a web server worker process terminates after
      handling N requests (omkd_max_requests) or if it grows beyond M
      megabytes of resident memory (omkd_max_memory)
  • New reorder protection feature for the handling of forwarded events
    this new feature (link to Deduplication and storm control in opEvents) was added to make the processing
    of forwarded events on a central server more robust
  • Automatically refreshing pages are now more resilient in how they react
    to unreliable client clocks or browser timer events.
  • Various bug fixes and robustness and efficiency improvements in both GUI and backend.
  • Planned outages as reported by NMIS are now translated into stateful events by opEvents.


opEvents 2.4.1

This maintenance release of opEvents was published on 2 Nov 2017.

...