Skip to end of metadata
Go to start of metadata
Details on alerts, whether adding new or editing existing.
Adding and editing of alerts is only active with the required privileges. See Establish User Role & Access Permissions.

Steps below begin from device's Facepage, click here for details on navigating to the resource to be removed.


Navigate to the Agent Instance

From the platform or instrument Facepage, scroll down to the bottom block with the Control Information tab. Activate the tab and click on the link to the Agent Instance.


Enable Editing of the Agent Instance

From the Agent Instance page, enable Edit from the main upper right drop down menu.


Add a New or Edit an Existing Alert

Edit an existing alert by adjusting the text in the alert box.

Use the 'X' button on the upper right of the alert box to delete the alert.

Click Add to create a new alert.

Alerts are described here.



Alert Example (LateDataAlert)

"stream_name" : "parsed",
"alert_type" : 2,
"name" : "late_data_warning",
"aggregate_type" : 1,
"alert_class" : "LateDataAlert",
"description" : "Expected data has not arrived.",
"time_delta" : 2

attribute name
description
stream_name the name of the stream which this alert is monitoring
alert_type is an enum:
StreamAlertType: !enum (values=(WARNING=1, ALARM=2, ALL_CLEAR=3), default=ALL_CLEAR))
where warning it yellow, alarm is red, and all_clear is green
name used in the event that is published when this alert goes active or is cleared
aggregate_type
an enum:
AggregateStatusType: !enum (values=(AGGREGATE_COMMS=1, AGGREGATE_DATA=2, AGGREGATE_LOCATION=3, AGGREGATE_POWER=4), default=AGGREGATE_COMMS))
tells the agent which of the four aggregate status' this alert applies to. Late data warning is applied to the comms status so if last_data_warning is yellow then comms status will be at least yellow. The aggregate status shows the highest level of any alert that is applied to it
NOTE:  The AggregateStatusType values correspond to the following status indicators on a resource's facepage: 
AGGREGATE_COMMS=1      =>   Communications    

AGGREGATE_DATA=2           =>   Data Flow              

AGGREGATE_LOCATION=3   =>   Location               

AGGREGATE_POWER=4        =>   Power                  
alert_class There are currently only a few alert classes.
  --  LateDataAlert - specifically to check for the delta between arriving granules.
  --  IntervalAlert    - checks that the value of a parameter on the stream is within a given range. Either lower_bound, upper_bound or both can be specified. You cannot specify two levels for one IntervalAlert (1-10 is green, 11-20 is yellow, >21 is red), this would be two different interval alerts.
  --  There are also a couple other types (StateAlert, CommandErrorAlert) RSNEventAlert was never fully integrated with the RSN team, it was intended to package RSN push alerts into CI events.
description used in the event that is published when this alert goes active or is cleared
time_delta specific to the alert_class LateDataAlert,
The unit is seconds and may be specified as an integer, or a float if fractional seconds are desired.
It is set to the max delta allowed between incoming granules

Stream alerts are either active or clear, once they are active then they will stay active until an action happens that causes the alert to clear. For late data alert, when a new granule arrives then the alert is set to all_clear. If this is a Interval alert checking that one of the parameters on the stream stays in a valid range then it will be set active when an out of range value is received then subsequently set to all_clear when a valid value is received.

With the delivery of R3 users will be able to upload code that can run in a data process which monitors a data stream and publishes events in whatever manner they prefer. That will provide an additional avenue to monitoring (for example, they are not restricted to the current set of alert classes) .



Alert Example (IntervalAlert)

"stream_name" : "status",
"alert_type" : 3,
"name" : "current_warning_interval",
"aggregate_type" : 4,
"alert_class" : "IntervalAlert",
"description" : "Current is above normal range.",
"value_id' : "port_current",
"lower_bound" : 10.5,
"lower_rel_op" : "<",
"upper_bound" : 15.0,
"upper_rel_op" : "<="

attribute name
description
value_id the name of the value in the stream which this alert is monitoring
lower_bound the lower limit for the value
lower_rel_op the operator (< or <=) to plug into the equation (lower_bound op value).  In the
above example this would become (10.5 < value).
upper_bound the upper limit for the value
upper_rel_op the operator (< or <=) to plug into the equation (value op upper_bound).  In the
above example this would become (value <= 15.0).

The lower and upper bounds and their operators satisfy the equation

(lower_bound lower_rel_op value upper_rel_op upper_bound). 

So in the example above that would become (10.5 < value <= 15.0)



Alert Example (StateAlert)

"alert_type" : 1,
"name" : "comms_warning",
"aggregate_type" : 1,
"alert_class" : "StateAlert",
"description" : "Detected comms failure.",
"alert_states" : ["RESOURCE_AGENT_STATE_LOST_CONNECTION",
                         "RESOURCE_AGENT_STATE_ACTIVE_UNKNOWN"],
"clear_states'": ["RESOURCE_AGENT_STATE_IDLE",
                         "RESOURCE_AGENT_STATE_COMMAND",
                         "RESOURCE_AGENT_STATE_STREAMING"]

attribute name
description
alert_states A list of states that cause the alert to be generated.
valid states are:
'RESOURCE_AGENT_STATE_POWERED_DOWN'
'RESOURCE_AGENT_STATE_UNINITIALIZED'
'RESOURCE_AGENT_STATE_INACTIVE'
'RESOURCE_AGENT_STATE_IDLE'
'RESOURCE_AGENT_STATE_STOPPED'
'RESOURCE_AGENT_STATE_COMMAND'
'RESOURCE_AGENT_STATE_STREAMING'
'RESOURCE_AGENT_STATE_TEST'
'RESOURCE_AGENT_STATE_CALIBRATE'
'RESOUCE_AGENT_STATE_DIRECT_ACCESS'
'RESOURCE_AGENT_STATE_BUSY'
'RESOURCE_AGENT_STATE_LOST_CONNECTION'
'RESOURCE_AGENT_STATE_ACTIVE_UNKNOWN'
clear_states
A list of states that cause the alert to be cleared.  The valid states are the same as for the alert_states attribute.



Alert Example (CommandErrorAlert)

"alert_type" : 1,
"name" : "comms_warning",
"aggregate_type" : 1,
"alert_class" : "CommandErrorAlert",
"description" : "Detected instrument failure.",
"command" : "RESOURCE_AGENT_EVENT_GO_ACTIVE",
"clear_states'": ["RESOURCE_AGENT_STATE_IDLE",
                         "RESOURCE_AGENT_STATE_COMMAND",
                         "RESOURCE_AGENT_STATE_STREAMING"]

attribute name
description
command The command that generates the alert if it fails.
clear_states
A list of states that cause the alert to be cleared. 
valid states are:
'RESOURCE_AGENT_STATE_POWERED_DOWN'
'RESOURCE_AGENT_STATE_UNINITIALIZED'
'RESOURCE_AGENT_STATE_INACTIVE'
'RESOURCE_AGENT_STATE_IDLE'
'RESOURCE_AGENT_STATE_STOPPED'
'RESOURCE_AGENT_STATE_COMMAND'
'RESOURCE_AGENT_STATE_STREAMING'
'RESOURCE_AGENT_STATE_TEST'
'RESOURCE_AGENT_STATE_CALIBRATE'
'RESOUCE_AGENT_STATE_DIRECT_ACCESS'
'RESOURCE_AGENT_STATE_BUSY'
'RESOURCE_AGENT_STATE_LOST_CONNECTION'
'RESOURCE_AGENT_STATE_ACTIVE_UNKNOWN'

Contact the helpdesk@oceanobservatories.org for more details on setting up appropriate alerts for operation of devices.


Labels

ci_user_doc ci_user_doc Delete
ci_adv_task ci_adv_task Delete
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.