AM2432: Profinet cyclic write problem after a reboot with S2 redundancy

Hi,

are there any updates?
I add that with a single PLC without redundancy MRP and S2, the problem doesn't occur, so I think it's specifically related to redundancy.

Thank you,

Best Regards,

Andrea

Hi,

are there any updates?
Is there any information that might be helpful?

Thank you,

Best Regards,

Andrea

Hi Andrea,

by "except for cyclic writes" , are you referring to L3 records write requests or L2 PPMs?

Can you help pointing out this faulty behavior with more details in the shared traces:   
- What's the expected behavior?
- What's the faulty behavior of the device?
- Which parts (lines) of the traces exactly reflect this discrepancy. 

Regards,
Assila

Hi Assila,

thank you for your reply.

Hi Andrea,

Thanks for the follow-up.

The  highlighted "IOPS" status is set to bad by controller, and the reason is the submodule plugged in slot 1/subslot 1 which seems to be wrong.

I checked all module/submodule parametrization sequences from the shared traces and could not see where the mod with ID:0x5 was plugged in slot 1.

- Where the shared traces captured at the same time from .51 and .52 ports?  
- Could you trigger a factory reset and re-run the test if you are storing the module/submodule config in the NVRAM?
- If possible, I would suggest you run same test with the ET200SP if it does support S2, just to confirm if this issue is showing for our device only.

Regards,
Assila

Hi Assila,

thank you for your reply.

Assila Belhouichet said:

The  highlighted "IOPS" status is set to bad by controller, and the reason is the submodule plugged in slot 1/subslot 1 which seems to be wrong.

I checked all module/submodule parametrization sequences from the shared traces and could not see where the mod with ID:0x5 was plugged in slot 1.

Could you tell me in which trace and which packet number suggests that the submodule plugged into slot 1/subslot 1 is incorrect and not accepted? Because, in fact, it shouldn't be module 0x5 but module 0x30.

Assila Belhouichet said:

- Where the shared traces captured at the same time from .51 and .52 ports?  

The traces were made just few seconds apart, I think a maximum of 4–5 seconds apart (just the time needed to start them on two different PCs).

Assila Belhouichet said:

- Could you trigger a factory reset and re-run the test if you are storing the module/submodule config in the NVRAM?

I don't save the latest configuration in flash memory (unlike what I think I saw in the demo). I can also perform a factory reset, but it doesn't delete anything related to the module configuration. In the defaultIoSubmodList I use I only enter DAP, PDEV Interface and PDev ports:

PN_API_IOD_SubmodListEntry_t defaultIoSubmodList[]
= {
   //api    slot    subslot     mod_id                  submod_id               in_data out_data    im0_support
   //                                                                           _len    _len
    {0,     0,      1,          PN_CFG_IDD_DAP_MOD_ID,  PN_CFG_IDD_DAP_SUB_ID,  0,      0,          PN_API_IOD_Im0Submodule + //DAP
                                                                                                    PN_API_IOD_Im0Device},
    {0,     0,      0x8000,     PN_CFG_IDD_DAP_MOD_ID,  PN_CFG_IDD_DAP_IF_ID,   0,      0,          PN_API_IOD_Im0Submodule}, //PDEV interface
    {0,     0,      0x8001,     PN_CFG_IDD_DAP_MOD_ID,  PN_CFG_IDD_DAP_P1_ID,   0,      0,          PN_API_IOD_Im0Submodule},
    //PDEV port1
    #if (PN_API_IOD_PDEV_NUM_OF_PORTS >= 2)
    {0,     0,      0x8002,     PN_CFG_IDD_DAP_MOD_ID,  PN_CFG_IDD_DAP_P2_ID,   0,      0,          PN_API_IOD_Im0Submodule},
    //PDEV port2
    #endif
    #if (PN_API_IOD_PDEV_NUM_OF_PORTS >= 3)
    {0,     0,      0x8003,     PN_CFG_IDD_DAP_MOD_ID, PN_CFG_IDD_DAP_P3_ID,    0,      0,          PN_API_eIm0Submodule},  //PDEV port3
    #endif
    #if (PN_API_IOD_PDEV_NUM_OF_PORTS >= 4)
    {0,     0,      0x8004,     PN_CFG_IDD_DAP_MOD_ID, PN_CFG_IDD_DAP_P4_ID,    0,      0,          PN_API_eIm0Submodule},  //PDEV port4
    #endif
    {0,     1,      1,          PN_CFG_IDM_MOD_IO_ID,   PN_CFG_IDS_IO_IF_ID,    0,      0,          (PN_API_IOD_Im0Submodule | //IO subslot interface
                                                                                                    PN_API_IOD_Im0Module)},
    {0,     1,      2,          PN_CFG_IDM_MOD_IO_ID,   PN_CFG_IDS_IO_T4_ID,    4,      4,          PN_API_IOD_Im0Submodule}, //IO subslot
    {0,     1,      3,          PN_CFG_IDM_MOD_IO_ID,   PN_CFG_IDS_IO_T2_ID,    2,      2,          PN_API_IOD_Im0Submodule}  //IO subslot
};

However, the problematic module is the one with ID 0x30, which is a module with the property FixedInSlots="1" in slot 1, subslot 1. Could that have something to do with it? 

Assila Belhouichet said:

- If possible, I would suggest you run same test with the ET200SP if it does support S2, just to confirm if this issue is showing for our device only.

I don't have a trace available, but it seems that ET200SP doesn't have the same problem.

Hi,

are there any updates?
Is there any information that might be helpful?

Thank you,

Best Regards,

Andrea

Hi,

are there any updates?
Is there any information that might be helpful?

Thank you,

Best Regards,

Andrea