Skip to content

Latest commit

 

History

History
173 lines (128 loc) · 6.67 KB

README_json.md

File metadata and controls

173 lines (128 loc) · 6.67 KB
Raw

JSON Runner

JSON.i

Important: when using run.sh, users should modify the corresponding .json.i file NOT the generated .json file. .json.i denotes incomplete or interim json file as it has fields intentionally left blank for run.sh to fill in automatically.

Acceptable inputs and range for json.i

Category Field Type Min Max
Mandatory/General publisher_interface String
Mandatory/General subscriber_interface String
Mandatory/General use_xdp Boolean
Mandatory/General packet_count Int 1 10000000
Mandatory/General cycle_time_ns Int 100000 5000000
Mandatory/General polling_duration_ns Int 0 10000000
Publisher url String
Publisher pub_id String
Publisher dataset_writer_id Int 0 99999
Publisher writer_group_id Int 0 99999
Publisher early_offset_ns Int 0 1000000
Publisher publish_offset_ns Int 0 10000000
Publisher socket_prio Int 0 7
Publisher two_way_data Boolean
Publisher cpu_affinity Int 0 3
Publisher xdp_queue Int 1 3
Subscriber url String
Subscriber sub_id String 0 99999
Subscriber subscribed_pub_id Int 0 99999
Subscriber subscribed_dataset_writer_id Int 0 99999
Subscriber subscribed_writer_group_id Int 0 99999
Subscriber offset_ns Int 0 10000000
Subscriber subscriber_output_file String
Subscriber temp_file_dir String
Subscriber two_way_data Boolean
Subscriber cpu_affinity Int 0 3
Subscriber xdp_queue Int 1 3

Acceptable inputs and range for tsn-json.i

Refer to examples provided by the project.

OPCUA-SERVER - AF_PACKET & AF_XDP OPCUA-based application

OPCUA-SERVER is a C-application that can transmit and receive ETH_UADP packets using AF_PACKET or AF_XDP sockets using libopen62541 (OPCUA-based library) APIs. It accepts only 1 .json.i file as input.

NOTE: The PLAT command line argument refers to either the CPU platform or the ethernet controller (e.g. the i225), see in json folder the available platforms.

About: 1-port

Very simple, just 1 pub/sub thread

At the end userspace-userpsace example, 1mil packet blabla

Usage: 1-port

Similar to TSQ & TXRX-TSN examples, 2 platforms are required to be connected to each other via a single-ethernet connection.

  1. [Board A & B] Build the project.

    cd /usr/share/iotg-tsn-ref-sw/
./build.sh

  2. [Board A] Run the setup script to initialize IP and MAC address, then start clock synchronization and setup TAPRIO + ETF qdiscs.

    cd /usr/share/iotg-tsn-ref-sw/
./run.sh $PLAT $IFACE opcua-pkt1a init
./run.sh $PLAT $IFACE opcua-pkt1a setup

  3. [Board B] Run the setup script to initialize IP and MAC address, then start clock synchronization and setup ingress qdisc.

    cd /usr/share/iotg-tsn-ref-sw/
./run.sh $PLAT $IFACE opcua-pkt1b init
./run.sh $PLAT $IFACE opcua-pkt1b setup

  4. [Board B] Start listening for packets.

    ./run.sh $PLAT $IFACE opcua-pkt1b run

  5. [Board A] Immediately after step 3, start transmitting packets. The application will terminate itself after completion. Note that AF_PACKET or opcua-pkt* configurations will take longer as they use much higher packet intervals compared to AF_XDP or opuca-xdp* configurations.

    ./run.sh $PLAT $IFACE opcua-pkt1a run

  6. [Board B] A gnuplot window should appear showing the time taken for each packet to traverse from the application in Board A to Board B.

    Each run will enumerate and store its raw timestamps/data & plot images in the results- folder for easy reference, repeat steps 3-4 as needed.

About: 2-port

In this example type, 2 platforms are required to be connected to each other via two ethernet connections. Meaning there is a total of four ethernet interfaces (as opposed to the usual two).

A simplified explanation of the data flow:

  1. Board A's first interface ($IFACE1) is used to transmit
  2. Board B's first interface ($IFACE1) is used to receive
  3. Board B's second interface ($IFACE2) is used to transmit back to Board A
  4. Board A's second interface ($IFACE2) is used to receive

Usage: 2-port

When using dual-ports, the $PLAT should be replaced with "ehl2". (TGL not supported)

  1. [Board A & B] Build the project.

    cd /usr/share/iotg-tsn-ref-sw/
./build.sh

  2. [Board A] Run the setup script to initialize IP and MAC address, then start clock synchronization and setup TAPRIO + ETF qdiscs.

    cd /usr/share/iotg-tsn-ref-sw/
./run.sh $PLAT $IFACE $IFACE2 opcua-pkt3a init
./run.sh $PLAT $IFACE $IFACE2 opcua-pkt3a setup

  3. [Board B] Run the setup script to initialize IP and MAC address, then start clock synchronization and setup ingress qdisc.

    cd /usr/share/iotg-tsn-ref-sw/
./run.sh $PLAT $IFACE $IFACE2 opcua-pkt3b init
./run.sh $PLAT $IFACE $IFACE2 opcua-pkt3b setup

  4. [Board B] Start listening for packets.

    ./run.sh $PLAT $IFACE $IFACE2 opcua-pkt3b run

  5. [Board A] Immediately after step 3, start transmitting packets.

    ./run.sh $PLAT $IFACE opcua-pkt3a run

  6. [Board A] A gnuplot window should appear showing the time taken for each packet to traverse from the application in Board A to Board B and back to Board A.

    Each run will enumerate and store its raw timestamps/data & plot images in the results- folder for easy reference, repeat steps 3-4 as needed.