This example shows how we can create the opEvents object for use inside a parser plugin.
...
Code Block | ||||
---|---|---|---|---|
| ||||
package Event_State_Example; our $VERSION="0.0.0"; use lib "/usr/local/omk/lib"; use strict; use OMK::Common; use OMK::opEvents; use OMK::Log; # arguments: the line (currently being parsed), # and reference to the live event properties # returns: (status-or-error) # # zero or undef: parsing for this event is aborted, # and no event is created. # 1: indicates success, event is created and changed event # properties are incorporated. # any other value: treated as error message, changed event # properties are NOT incorporated but event parsing continues. sub parse_enrich { my ($line, $event) = @_; my $confCommon = loadOmkConfTable(conf=> "opCommon", dir=> "/usr/local/omk/conf"); my $logger = OMK::Log->new(level => $confCommon->{"omkd_log_level"} || 'info', path => $confCommon->{'<omk_logs>'}."/opEvents.log"); my $OPE = OMK::opEvents->new(config => $confCommon, logprefix => "Plugin::Event_State_Example", log => $logger); $OPE->getDb(); $event->{Plugin_Used} = "Event_State_Example"; return 1; } |
getEventLogsModel
# args: at least log_name (=db collection, log_name can be: events, rawLogs, logArchive, or actionlog)
# and time_start/end AND/OR a set of any
# of id, node_uuid or node_name,
# type/element/details/action/archive/entry/acknowledged/escalate/priority/event_id
# to select events
#
# arg sort: mongo sort criteria
# arg limit: return only N records at the most
# arg skip: skip N records at the beginning. index N in the result set is at 0 in the response
# arg paginate: sets the pagination mode, in which case the result array is fudged up sparsely to
# return 'complete' result elements without limit! - a dummy element is inserted at the 'complete' end,
# but only 0..limit are populated
In this example we are getting an event by its ID
...
Code Block | ||
---|---|---|
| ||
'_id' => $arg{id}, 'time' => { '$gte' => $time_start, '$lt' => $time_end }, 'event' => $arg{event}, 'node_uuid' => $arg{node_uuid}, 'type' => $arg{type}, 'element' => $arg{element}, 'details' => $arg{details}, 'eventid' => $arg{event_id}, # only useful in actionlog 'action' => $arg{action}, # only useful in actionlog 'archive' => $arg{archive}, # only useful in archive log 'entry' => $arg{entry}, # only in raw log 'state' => $arg{state}, 'nodeinfo.configuration.location' => {'$regex' => $arg{'nodeinfo.configuration_location'} || $arg{location}}, 'nodeinfo.configuration.group' => {'$regex' => $arg{'nodeinfo.configuration_group'} || $arg{group}}, 'acknowledged' => numify($arg{acknowledged}), 'escalate' => numify($arg{escalate}), 'priority' => numify($arg{priority}), }); |
Updating Events
# updates an event with the given details
# args: _id (for finding the event), _constraints (to disable db key munging),
# everyting else is set as record content, as-is - except "status_history", "trigger_eventids"
#
# status_history: optional but special: must be array and this array will be ADDED to an
# existing status_history array!
# trigger_eventids: always saved as array, and a new value is ADDED!
# buttons: always saved as array, and a new value is ADDED!
# returns undef if ok, error message otherwise (also logged)
Acknowledging an event
We can acknowledge an event by setting acknowledged => 1, and give it status history so we know who and when triggered the event to be acknowledged
...