And there is one more argument: 

We have a Python expert in our development team who wrote already the Python-to-C bindings. That means when running a Python 
script, we can already start/stop runs, write/read to the ODB etc. We only have to get the single stepping going which seems feasible to 
me, since there are some libraries like inspect.currentframe() and traceback.extract_stack(). For single-stepping there are debug APIs 
like debugpy. With Lua we really would have to start from scratch.

Stefan

I had to change the ODBXX API: https://bitbucket.org/tmidas/midas/commits/273c4e434795453c0c6bceb46bac9a0d2d27db18

The old C API is case-insensitive, meaning db_find_key("name") returns a key "name" or "Name" or "NAME". We can discuss if this is good or bad, but that's how it is since 30 years.

I now realized the the ODBXX API keys is case sensitive, so a o["NAME"] does not return any key "name". Rather, it tries to create a new key which of course fails. I changed therefore
the ODBXX to become case-insensitive like the old C API.

Stefan

Indeed this was missing from the very beginning. I added it, please report back if it's not working.

Stefan

I would be in favor of not curing the symptoms, but fixing the cause of the problem. I guess you put the watchdog disable into mhist, right? Usually mhist is called locally, so no mserver should be 
involved. If not, I would prefer to propagate the watchdog disable to the mserver side as well, if that's not been done already. Actually I never would disable the watchdog, but set it to a reasonable 
maximal value, like a few minutes or so. In that case, the client gets still removed if it crashes for some reason.

My five cents,
Stefan 

I would handle this actually like symbolic links are handled under linux. If you delete a symbolic link, the link gets 
detected and NOT the file the link is pointing to.

So I conclude that the "follow links" is a misconception and should be removed.

Stefan

> Dear experts,
>       in the CYGNO experiment, we readout CMOS cameras for optical readout of GEM-TPCs. So far, we only developed the readout for a single camera. In the future, we will have multiple cameras to read out (up to 18 in the next phase of the experiment), and we are investigating how to optimize the readout by exploiting parallelization.
> 
> One idea is to start parallel threads within a single equipment. Alternatively, one could associate different equipment with each camera and run an Event Builder. Perhaps other solutions did not come to mind. Which one would you regard as the most effective and elegant?
> 
> Thank you very much,
>            Francesco

In principle both will work. It's kind of matter of taste. In the multi-threaded approach one has a single frontend to start and stop, and in the second case you have to start 18 individual frontends and make sure that they are running. 

For the multi-threaded frontend you have to ensure proper synchronization between the threads (like common run start/stop), and in the end you also have to do some event building, sending all 18 streams into a single buffer. As you know, multi-thread programming can be a bit of an art, using mutexes or semaphores, but it can be more flexible as the event builder which is a given piece of software.

Best,
Stefan

> Hello, I'm currently writing a paper about making a dataset publicly available 
> from the University of Minnesota and it's a MIDAS dataset.
> 
> I'd like to use an image that shows the MIDAS data format (that's been slightly 
> modified to fit in the paper) and am wondering (1) if I could get permission to do 
> so and (2) what the preferred attribution would be. 

Feel free to use whatever you like, the documentation is on the same open source license as midas itself.

Stefan

Today the way the history system is configured has changed. Whenever one adds new equipment 
variables, the logger has to add that variable to the history system. Previously, this happened during 
startup of the logger and at run start. We have now a case in the Mu3e experiment where we have many 
variables and the history configuration takes about 15 seconds, which delays data taking considerably.

After discussion with KO we decided to remove the history re-configuration at run start. This speeds up 
the run start considerably, also for other experiments with many history variables. It means however that 
once you add/remove/rename any equipment variable going into the history system, you have to restart the 
logger for this to become active.

https://bitbucket.org/tmidas/midas/commits/c0a14c2d0166feb6b38c645947f2c5e0bef013d5

Stefan

We had issues with the frontend write cache size and the way it was implemented in the frontend 
framework mfe. Specifically, we had two equipments like in the experiment/examples/frontend.cxx, one 
trigger event and one periodic event. While the trigger event sets the cache size correctly in the 
equipment table, the periodic event (defined afterwards) overwrites the cache size to zero, causing slow 
data taking. 

The underlying problem is that one event buffer (usually "SYSTEM") can only have one cache size for all 
events in one frontend. To avoid mishaps like that, I remove the write cache size from the equipment table 
and instead defined a function which explicitly sets the cache size for a buffer. In your frontend_init() you 
can now call

  set_cache_size("SYSTEM", 10000000);

to set the cache size of the SYSTEM buffer. Note that 10 MiB is the minimum cache size. Especially for 
smaller events, this dramatically can speed up your maximal DAQ rate.

Full commit is here:

  https://bitbucket.org/tmidas/midas/commits/24fbf2c02037ae5f7db74d0cadab23cd4bfe3b13

Note that this only affects frontends using the mfe framework, NOT for those using the tmfe framework.

Stefan

I added now code to the logger so it properly complains if there would be a leading space in a variable name.

Stefan

> By accident we had history variables with leading spaces. The history schema check then decides that this is a new variable (the leading space is not read from the history file) and starts a new file. We found this because the run start became slow due to the many, many history files created. It would be nice to just get an error if one has a malformed variable name like this.
> 
> How to reproduce: Try to put a variable with a leading space in the name into the history, repeatedly start runs.
> Sugested fix: Produce an error if a history variable has a leading space.

Since this memory leak is quite obvious, I pushed the fix to develop.

Stefan

Hello!

My name is Soichiro Kuribayashi and I am a Ph.D. student at Kyoto University. 
I'm a T2K collaborator and working for Super FGD which is new detector in ND280.

I'm a beginner of MIDAS and I've just started to develop the DAQ software with 
MIDAS for Super FGD.
For the DAQ of Super FGD, we will run remotely front end part of MIDAS on ZYNQ 
which is system on chip.

For this remote control of front end part with mserver, we have to mount home 
directory of DAQ PC(Cent OS8) on that of Linux on ZYNQ.
So I wonder if we should use NFS(Network file system) + NIS(Network information 
service) + autofs for the mounting. Is it correct?

If you have any information or any suggestion for the remote control on chip, 
please let me know.

Best regards,
Soichiro 

Dear Konstantin,

Thank you very much for your reply and detailed information.
I would appreciate if you could help us.

> I can also suggest that on your frontend SOC/FPGA machine, you boot linux
> using the "nfs-root" method. This way, the local flash memory only
> contains a boot loader (and maybe the linux kernel image, depending on
> bootloader limitations). The rest of the linux rootfs can be on your
> central development machine. This way management of flash cards,
> confusion with different contents of local flash and need to make backups
> of frontend machines is much reduced.

As you said, we can run complete Linux (Ubuntu 16) on ZYNQ and I'm using common NFS 
system now. However, I didn't know "nfs-root" method which you mentioned and this method 
seems to be reasonable way to just share linux rootfs.
First of all, I will try this method for simpler system.

> If you use a fast SSD and ZFS with deduplication, you will also have good
> performance gain (NFS over 1gige network to server with fast SSD works
> so much better compared to the very slow SD/MMC/NAND flash).
>
> I can point you to some of my documentation how we do this.

I'm concerned about such performance and I have checked the performance with common NFS 
over gige network and my DAQ PC roughly(data transfer rate ~ O(10) MByte/sec). However, I 
didn't know the ZFS and also how we can have performance gain with a fast SSD and ZFS.
Please let me know your documentation how to do it if possible.

> I would say you are on a good track. For early development on just one board,
> pretty much any way you do it will work, but once you start scaling up
> beyound 3-4-5 frontends, you will start seeing benefits from common NFS-mounted
> home directories, NFS-root booted linux, etc.

I'm developing with just one board and common NFS-mounted now. I'm looking forward to 
seeing such benefits when I will use multiple frontends.
 
> And of course you may want to study the existing ND280/FGD DAQ. I hope you
> have access to the running system at Jparc. If not, I have a copy of
> pretty much everything (except for running hardware, it is stored in the basement, 
> dead) and I can give you access.

I don't have access to the system at Jparc, but Nick has told us where FGD DAQ code is.
Is bellow URL everything of code of FGD DAQ?
https://git.t2k.org/hastings/fgddaq/-/tree/master

Best regards,
Soichiro

Dear Stefan,

Thank you very much for your help.

I have already contacted someone who has used ZYNQ in that order and It's working fine for now.
But, I'll let you know if something goes wrong.

Best regards,
Soichiro