Doca-flow-inspector service can't start properly on BF-2

I am trying to deploy doca-flow-inspector service, but the container consistently can’t start properly, I see the following error in the container log, how to solve it, is it because my hardware doesn’t support it?

2023-06-13T02:43:59.194063355Z stderr F mlx5_net: [mlx5dr_action_create_generic]: Cannot create HWS action since HWS is not supported
2023-06-13T02:43:59.194395224Z stderr F [02:43:59:194237][DOCA][ERR][dpdk_flow_hws_legacy:143]: failed to configure flow hws port 0 - rte flow configure, type 1 message: fail to configure port
2023-06-13T02:43:59.194438251Z stderr F [02:43:59:194311][DOCA][ERR][dpdk_engine:1694]: failed to start port 0 - init port, ret=-1
2023-06-13T02:43:59.194513879Z stderr F [02:43:59:194383][DOCA][INF][utils_hash_table:151]: hash table destroyed
2023-06-13T02:43:59.194654076Z stderr F [02:43:59:194510][DOCA][INF][utils_hash_table:151]: hash table destroyed
2023-06-13T02:43:59.19470789Z stderr F [02:43:59:194567][DOCA][INF][utils_hash_table:151]: hash table destroyed
2023-06-13T02:43:59.19507929Z stderr F [02:43:59:194983][DOCA][INF][utils_hash_table:151]: hash table destroyed
2023-06-13T02:43:59.195093138Z stderr F [02:43:59:195016][DOCA][INF][dpdk_port:230]: Dpdk port 0 destroyed successfully with 8 queues
2023-06-13T02:43:59.195163799Z stderr F [02:43:59:195039][DOCA][ERR][INSPEC:CORE:1031]: Failed to initialize doca flow port: dpdk port start failed
2023-06-13T02:43:59.195256229Z stderr F [02:43:59:195067][DOCA][INF][doca_flow_layer:478]: Doca flow layer destroyed

Many thanks in advance,


hi goodtuninger

You miss some config for enable HWS.
please reference :

14.2. Running the Sample

CLI example for running the samples with “vnf,hws” mode:
./build/doca_<sample_name> -a auxiliary:mlx5_core.sf.2,dv_flow_en=2 -a auxiliary:mlx5_core.sf.3,dv_flow_en=2 – -l 60

For hws(hardware steering) detail:

dv_flow_en parameter [int]

Value 2 enables the WQE based hardware steering. In this mode, only queue-based flow management is supported.

Thank you
Meng, Shi