Configuration files (including instrument agent driver startup)
Cabled services are mainly managed on the Portland (West Coast) servers. The provide a number of utilities to control and manage data from the cabled arrays (Regional Scale Nodes and Coastal Endurance).
Services are configured using the supervisor utility. The configuration can usually be found on the listed machine under the ~/supervisors directory. A configuration file provides the details for managing the services (e.g. pa1.conf).
Initial configuration of supervisor requires running the supervisor daemon which must be run under the port_agent environment:
/home/asadev/anaconda/envs/port_agent/bin/python /home/asadev/anaconda/envs/port_agent/bin/supervisord -c pa2.confAfter supervisord is running on the system, supervisorctl can be used to control the services.
For example, to configure the services on portland-01:
cd ~/supervisors/pa1
source activate instruments
supervisorctl -c pa1.conf
> start pa_monitor
> quitSupervisor processes are run on the following machines using the given configurations:
| location | configuration | purpose |
|---|---|---|
| portland-01 | pa1/pa1.conf | port agents |
| portland-02 | pa2/pa2.conf | port agents |
| portland-03 | oms_aa_server/oms_aa_svr.conf | Alerts and Alarms |
| portland-03 | ia/services.conf | InstrumentAgent & MissionExecution |
| portland-04 | drivers/pad_drivers.conf | platform engineering streams |
| portland-05 | drivers/drivers.conf | instrument drivers |
| portland-06 | drivers/drivers.conf | instrument drivers - backup |
| antelope | port_agent_supervisor/antelope.conf | port agents |
The services run by supervisor are listed below:
| name | service | purpose | location |
|---|---|---|---|
| reference designator | port_agent | communication link between specific instrument and its corresponding instrument agent driver | portland-01 portland-02 antelope |
| reference designator | run_driver | instrument driver | portland-05 portland-06 |
| events_shovel | shovel | portland-01 | |
| pa_monitor | ooi_port_agent.tools.monitor | email notification service if network connectivity is lost to instruments | portland-01 portland-02 |
| oms_alerts_shovel | shovel | forwards OMS alerts/alarms to production server | portland-01 |
| oms_extractor | creates engineering particles from engineering ports | portland-04 | |
| omsaasvr | oms_alert_alarms_server | creates alert/alarms from OMS events | portland-03 |
| pa_stats_shovel | shovel | forwards statistics of port agent traffic to production server | portland-01 portland-02 |
| particles_shovel | shovel | forwards instrument packets to production server | portland-01 |
Port agent files control the flow of data between the instruments and the instrument agent drivers. This allows for both streaming data from the instrument into the system as well as command and control of the instruments by the driver. Each instrument has a dedicated port agent. The reference designator is used to identify the port agent in the supervisor utility. For example, stop RS01SBPS-PC01A-07-CAMDSC102 will stop the port agent for the digital camera. Note that the VADCP and the MASSP require additional port agents; RS01SBPS-PC01A-06-VADCPA101_5 manages the interface for the 5th beam, -MCU -TURBO -RGA manage the controls for the individual units on the MASSP. Port agents are distributed across two servers. The port agent is configured with the instrument address and reference designator. For example the CTDBPO108 is started with the following command:
port_agent rsn 10.33.12.6 2101 2102 CE04OSBP-LJ01C-06-CTDBPO108
The run_driver script configures and runs an instrument agent driver that manages the connection to a specific instrument. The instrument agent driver provides parsing of the live instrument feed as well as command and control (C2) of the instrument. C2 is currently disabled. The driver configuration includes the driver code, reference designator, events and particles queue and configuration. For example: NUTNRA102 is configured with the following command:
run_driver mi.instrument.satlantic.suna_deep.ooicore.driver InstrumentDriver CE04OSPS-SF01B-4A-NUTNRA102 amqp://ooi-production:[email protected]/production?queue=ha.instrument_events amqp://ooi-production:[email protected]/production?queue=ha.instrument_particles /home/asadev/ooi-config-master/CE04OSPS-SF01B-4A-NUTNRA102.yaml
Startup requires a YAML configuration file that can be found in ooi-config/instrument_driver_startup. Currently the yaml files are installed at portland-05:~/ooi-config-master (a link to ~/ooi-config/instrument_driver_startup).
Forwards instrument events to uFrame. Executes the following command:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" ha.instrument_events ha.instrument_events qpid://ooiufs03.ooi.rutgers.edu Ingest.instrument_events
The port agent monitor provides email notification service in the event that network connectivity is lost to an instrument. The monitor.conf file is use to identify the email recipients and which connections are to be monitored. This service checks connectivity using ping hourly and only notifies when there is a failure. Executes the following command:
python -m ooi_port_agent.tools.monitor monitor.conf
Forwards OMS alerts (generated by oms_extractor) to production alert ingest path:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" ha.oms_alerts ha.instrument_events qpid://ooiufs03.ooi.rutgers.edu Ingest.instrument_oms_events
Generates engineering particles from the platform engineering elements. Executes the following command:
oms_extractor oms_extract_config.yml
The YAML file references the node configuration files which are used to specify the mapping of each engineering node to the corresponding OMS endpoint. These files are managed in the mi-instrument project, c.f. https://github.com/oceanobservatories/mi-instrument/tree/master/mi/platform/rsn/node_config_files
Converts OMS messages into OOI alert/alarms. Executes the following command:
oms_alert_alarm_server.py omsaasvr_p1.cfg
Forwards instrument particle statistics (kB/sec) to uFrame. Executes the following command:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" port_agent_stats port_agent_stats qpid://uft20.ooi.rutgers.edu port_agent_stats
Ingests live instrument data particles into OOI. Executes the following command:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" ha.instrument_particles ha.instrument_particles qpid://ooiufs03.ooi.rutgers.edu Ingest.instrument_particles
| name | service | purpose | location |
|---|---|---|---|
| particles_shovel_uft4 | shovel | forwards instrument particles to test server (uframe-4-test) | portland-01 |
| particles_shovel_uft20 | shovel | forwards instrument particles to test server (uft20) | portland-02 |
| uframe_test_shovel | shovel | forwards instrument particles to test server (uft20) | portland-02 |
| uframe_2_test_shovel | shovel | forwards instrument particles to test server (uframe-2-test) | portland-02 |
| uframe_3_test_shovel | shovel | forwards instrument particles to test server (uframe-3-test) | portland-02 |
Test for ingesting live cabled instrument data to the test machine, uframe-4-test. Executes the following command:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" uft4_instrument_particles ha.instrument_particles qpid://uframe-4-test.ooi.rutgers.edu Ingest.instrument_particles
Test for forwarding live cabled messages to the test machine, uft20 which may no longer be operational.
Test for broadcasting alerts to test UI. Executes the following command:
shovel "amqp://ooi-production:T1cZp68WanMZ53dQ@localhost/production" ha.oms_alerts ha.instrument_events qpid://ooiufs03.ooi.rutgers.edu Ingest.instrument_oms_events