GE Embedded Field Agent, Руководство пользователя

Страница: 80 / 87

GFK-2993F 80
For public disclosure
6.2.4 Configure Sending Data to the Predix Cloud
Predix Machine uses the Hoover Spillway and Data River bundles to select available data and forward it to a data destination
such as a Time Series Database. The Hoover Spillway must be configured to specify which of the available Machine Adapter
data subscriptions should be consumed, and sent to which Data River. To send the data to a Time Series Database, the WebSocket
River can be used. The WebSocket River must be configured to specify which Time Series Database instance the data should
be sent to. This is done by specifying a Zone ID, which is unique to each EdgeManager tenancy.
The steps to configure the Hoover Spillway and WebSocket River are:
1. Open the file com.ge.dspmicro.hoover.spillway-0.config for editing.
2. Edit line 60 to set com.ge.dspmicro.hoover.spillway.dataSubscriptions to include only the data subscription names
defined in section Configure the Field Agent Data Source .
Note When Predix Machine first starts, an error message may be logged for each configured subscription stating that the
subscription could not be found. This is because the Hoover Spillway can load prior to Machine Adapters, and will look for its
configured subscriptions before the subscriptions are available from configured Machine Adapters. Once the subscriptions are
available from the Machine Adapter(s), Hoover Spillway automatically re-attempts to connect to the configured subscriptions.
3. Verify the Hoover Spillway is configured to send the subscription data to the WebSocket River. The “WS
Sender Service” name comes from the WebSocket River name that was configured in the
com.ge.dspmicro.websocketriver.send-0.config file.
•
com.ge.dspmicro.hoover.spillway.destination=”WS Sender Service”
4. Verify the Hoover Spillway is configured to enable the Store and Forward service, which buffers data
while the connection to the Predix Cloud is broken. If the Store and Forward service is not desired, this
variable should be set to two double quotes (“”).
•
com.ge.dspmicro.hoover.spillway.storeforward="DefaultStoreForward"
5. Save and close com.ge.dspmicro.hoover.spillway-0.config .
6. Open com.ge.dspmicro.websocketriver.send-0.config for editing.
7. Verify line 54 is set to include the URL of the Time Series Database.
•
com. ge.dspmicro.websocketriver.send.destination.url="wss://gateway-predix-data-
services.run.aws-usw02-pr.ice.predix.io/v1/stream/messages"
8. Edit line 60 to include the Zone ID of the Time Series Database (provided upon activation of Time Series
Service).
•
com.ge.dspmicro.websocketriver.send.header.zone.value="xxxxxxxx-xxxx-xxxx-xxxx-
xxxxxxxxxxxx"
Note It is not recommended to disable Gzip compression for data sent to Predix Time Series Database. Disabling Gzip
compression will require additional upstream bandwidth and may decrease maximum number of published variables.
9. Save and close com.ge.dspmicro.websocketriver.send-0.config .
Note
All Machine Adapter timestamps are sent to and stored in the Predix Time Series Database as epoch time (the number
of seconds that have elapsed since January 1, 1970 at midnight UTC not counting leap seconds).
Note
Additional information on configuring the Hoover Spillway can be found by searching for “Configuring the Hoover
Service” in https://docs.predix.io .
Note
Additional information on configuring the WebSocket River can be found by searching for “Configuring the
WebSocket River” in https://docs.predix.io .
Note
Additional information on the Predix Time Series Database can be found by searching for “Time Series Service
Setup” in https://docs.predix.io .