Communication with AMCs via IPMB-L

1

Dear experts,

I am from the ATLAS LAr group and trying to understand the issue we have for the reset of IPMC at daily level. The newest observation showed there is a lost of communication with AMCs so I would like to know how CERN IPMC handling communication via IPMB-L and is there detailed documentation or code available?

Thanks,
Yi-Lin

2

Dear Yi-Lin,

Thanks for your message. Can you please tell me which board you are working on, and give me a bit more details about the problem you are encountering? Are you seeing “a lot of communications” or a “loss of communication”. That was not clear to me. What are you trying to achieve?

Cheers,
Ralf

3

Dear Ralf,

Sorry about the unclear description.

We are working on migrating LAPP IPMC to CERN IPMC for our trigger system. One carrier board has a CERN IPMC and four AMCs. We observed the CERN IPMC is restarting daily through checking the uptime of IPMC so we set the SOL session to log the events.

Before IPMC reset, we saw a lot of

iface_ipmb_send_message_ipmb: i2c NACK
: IPMB-L error: IPMC 7E: no acknowledge

5 lines of NACK and one line of IPMB-L error is for each output, and it repeats for thousands of lines.

Since we don’t see this kind of behavior in the LAPP IPMC, I am wondering if there is more document descripting IPMB-L part so that we can tune the parameters to achieve the stability. I checked the git repository and website and I couldn’t find more information for this.

Cheers,
Yi-Lin

4

Dear Yi-Lin,

Thanks for your reply.

What do you mean by the IPMC’s uptime; it does not have that, at least not directly. Can you tell me how you are seeing that it has rebooted?

The AMC IPMB-L NACK are, of course, serious, and could be pointing to a hardware problem. Are you sure the four AMCs are up and running? What does the address 7E correspond to?

Of course, we could give you access to the source code if your institute has signed the NDA. Can you please tell what institute you are working and who is the IPMC contact at your institute, if you know that.

Cheers,
Ralf

5

Dear Ralf,

Thank you for the follow-ups.

  • About the uptime, we use the TIMER_CALLBACK to count the seconds and get the uptime reading the variable.
  • We are sure the AMCs were running, We saw the the flood of NACK 5 hours after we inserted all the AMCs so the communication in the beginning was fine. I believe the 7E is address of one AMC we inserted.
  • I understand the issue of NDA. Unfortunately, the people working in our group don’t have the assignment of NDA with Pigeon Point. This is also my question. Is there extract information, which can be disclosed from your side under NDA? Ex: the parameters for timeout or sequentially checking AMCs. ( I don’t actually know what could be the details. I just want to seek any possibility we can tune anything for this since the document is very limited.)

Cheers,
Yi-Lin

6

Hi Yi-Lin,

If you tell me what your institute is then we can check if there is already an NDA registered, or we could register one.

And you told me already that you are working for ATLAS LAr. Is it for one particular board? COudl you please provide that information?

Cheers,
Ralf

7

Hi Ralf,

I am from the ICEPP, University of Tokyo and also in the Liquid Argon calorimeter group at ATLAS experiment. Please let me know if you need more information.

Cheers,
Yi-Lin

8

Hi Yi-Lin,

Thanks for your reply.

There is no NDA for ICEPP, Univeristy of Tokyo, yet. So, if you are still interested getting access to the source code of the CERN IPMC, could you please send me the email address of a responsible from your institute (a professor, or permanent staff member) so that we can send them the NDA contract for signature?

In the meantime, or alternatively, we were wondering if you could give us access to your repository for the IPMC software that you are building to run on the CERN IPMC. That way, we could have a look and see if we can spot any problematic statements.

Cheers,
Ralf

9

Hi Ralf,

Thank you for checking the NDA. I think I will stop here since this project is not the main project in our institute but I am following the development.

We copied files to the git repository for our group. Here is zip file for the code we are using.

We run the command for the image production.
python compile.py -c config_LARC.xml -b BOARD_LARC

Please let me know if you cannot access the file.

Cheers,
Yi-Lin