CPV missing TF handling #6483
CPV missing TF handling #6483
Conversation
sevdokim
requested review from
kharlov,
peressounko,
sawenzel and
shahor02
as code owners
shahor02
left a comment
shahor02
left a comment
There was a problem hiding this comment.
@sevdokim thanks, see a few comments below!
The formatting needs to be fixed. In general, it is better to avoid pushing so many unrelated changes in a single commit.
Since the FullCI will likely fail due to internal reasons and you are touching raw data handling, could you run yourself the FST (you need to have the AEGIS installed):
$O2_ROOT/prodtests/full_system_test.sh ?
| #pragma link C++ class o2::ctf::EncodedBlocks < o2::cpv::CTFHeader, 7, uint32_t> + ; | ||
|
|
||
| #pragma link C++ class o2::cpv::BadChannelMap + ; | ||
| #pragma link C++ class o2::TObjectWrapper < o2::cpv::BadChannelMap> + ; |
There was a problem hiding this comment.
TObjectWrapper is obsolete, please remove all these entries.
There was a problem hiding this comment.
Ok, thanks. Removed.
| /*enum RawErrorType_t { | ||
| kOK, ///< NoError | ||
| kOK_NO_PAYLOAD, ///< No payload per ddl (not error) | ||
| kRDH_DECODING, | ||
| kRDH_INVALID, | ||
| kNOT_CPV_RDH, | ||
| kSTOPBIT_NOTFOUND, | ||
| kPAGE_NOTFOUND, | ||
| kPAYLOAD_INCOMPLETE, | ||
| kNO_CPVHEADER, | ||
| kNO_CPVTRAILER, | ||
| kCPVHEADER_INVALID, | ||
| kCPVTRAILER_INVALID, | ||
| kSEGMENT_HEADER_ERROR, | ||
| kROW_HEADER_ERROR, | ||
| kEOE_HEADER_ERROR, | ||
| kPADERROR, | ||
| kUNKNOWN_WORD, | ||
| kPadAddress | ||
| };*/ |
| // cfgc.options().get<bool>("pedestal"), | ||
| // cfgc.options().get<std::string>("ccdb-url"), |
There was a problem hiding this comment.
why do you need to add options which are not needed?
There was a problem hiding this comment.
hello @shahor02
they are not needed. I removed them.
There was a problem hiding this comment.
Dear @shahor02 ,
I'll try to run full system test and let you know the result.
I also want to be sure that protection against missing TFs is working.
We can simulate missing TFs by command
o2-raw-file-reader-workflow --input-conf raw/CPV/CPVraw.cfg --loop 100 --drop-tf "CPV,3"
But what I need to do next to see if everything is working as expected?
Thank you & regards,
Sergey
There was a problem hiding this comment.
Hi @sevdokim
you should see in every 3d TF the message you put: "Sending empty output due to data type input with 0xDEADBEEF"
There was a problem hiding this comment.
Dear @shahor02 ,
I tried to run FST using following environment on ubuntu20.04:
alienv enter O2/latest AEGIS/latest
Unfortunately the test was failed at 2nd task:
Launching task: o2-sim --seed -1 -n 10 --configKeyValues Diamond.width[2]=6. -g pythia8hi -e TGeant4 -j 4 &> sim.log &
command o2-sim --seed -1 -n 10 --configKeyValues Diamond.width[2]=6. -g pythia8hi -e TGeant4 -j 4 had nonzero exit code 1
From the sim.log I see following:
Spawning particle server on PID 1270555; Redirect output to o2sim_serverlog
[INFO] Process 1270555 EXITED WITH CODE 0 SIGNALED 1 SIGNAL 6
from the o2sim_serverlog:
PYTHIA Error in BeamRemnants::setKinematics: kinematics construction failed
[ESC[01;36m17:11:29ESC[0m][ESC[01;33mWARNESC[0m] Particle to be tracked is not defined in PDG: pdg = 1000200489
[ESC[01;36m17:11:29ESC[0m][ESC[01;33mWARNESC[0m] Particle to be tracked is not defined in PDG: pdg = 1000180489
I suppose the error is not related to CPV but to non-recognition of some ions by the VMC?
I bet that my setup is not well defined and it was a cause of this failure. Any tips how to proceed?
Thank you!
Sergey
There was a problem hiding this comment.
Hi @sevdokim
Will run myself to check, we recently switched to G4, might be related to this.
There was a problem hiding this comment.
hi @shahor02 ,
done from my side! thanks!
|
@sevdokim The FST seems to be broken at the moment, I've tested your branch on existing raw data and had to apply some changes to make lost TF recovery to work. Please merge the sevdokim#2 I opened on your repo, then I will merge this PR w/o waiting again for the tests. |
RAWDATA Lifetime must be Optional, add extra options
shahor02
left a comment
shahor02
left a comment
There was a problem hiding this comment.
Merging as the tests were passed before and last fix was tested locally.
* CPV missing TF handling * clang-formatted commit * format fix of CMakeLists.txt * fix for cpv macros * fix for macros/CMakeLists.txt * RAWDATA Lifetime must be Optional, add extra options Co-authored-by: sevdokim <sergey.evdokimov@cern.ch> Co-authored-by: shahoian <ruben.shahoyan@cern.ch>
2 participants
Hello!
In this PR missing TF protection was implemented to RawToDigitConverter.
Also I moved Hit.cxx, CalibParams.cxx, BadChannelMap.cxx, and Pedestals.cxx to DataFormats/Detectors/CPV/.
Also I added configuration of CCDB Url to RawToDigitConverter and RawWriter.