This page contain the possible issue with LTDBs control system that need expert intervention and how to recover it. Please feel free to complete it -- EtienneMarieFortin - 2021-06-15

Restart detector after power cut :

NEW Methode

1) All this step have to be done on the 8 pc-lar-felix-ltdb-XX

• kill the 3 servers (Felix, OpcUaSca, OpcUaLarLtdb)
• start Felix
  • configure
  • start
• start OpcUaSca
• start OpcUaLarLtdb

2) Then

• configure ltdb via GUI interface

3) if there are red LTDB boxes,
then click on this box and redo the configuration steps for this LTDB in the open window
if still in error, click, in this window, on the concerned box and in the panel "monitoring" check if errors on sca or GBTx,

=> if yes, click on the reset buttons (a little bit all, it ends up working) if the issue is still there===> check logs files (OpcUaSca and OpcUaLtdb.....and do the"obsolete methode " that follow

4) Then

• do clkdes

OBSOLETE Methode but always active:

all this step have to be done on the 8 pc-lar-felix-ltdb-XX

• kill the 3 servers (Felix, OpcUaSca, OpcUaLarLtdb)
• start Felix
  • configure
  • start
• start OpcUaSca
  • check log
    • if you see error this mean you have some sca not online so they need to be reset
      • for that go on tools/board_reset and use the reset with the elink (you see this number in the log) or with the position of the ltdb.

The position of the ltdb is in the OpcUaSca configuration file for exemple ltdb12_scafelix2 => board_reset> ./reset_from_from_ltdb_position.sh 12 Attention, dans la BdD de Clément au delà de position 8, la position est le numero +1

• start OpcUaLarLtdb
• configure ltdb
• do clkdes

Start project on PCATLAR05 with cronjob :

• cd /det/dcs/Production/ATLAS_DCS_LAR/cron
  • sudo -u atlardcs crontab pcatllar05.cronfile
  • sudo -u atlardcs crontab -l to see all project running

Issue during the restart of OpcUaSca with new config :

• Check if difference existe between tne old config and the new
  • Diff config_wendy_“old”.xml config_wendy_“new”.xml
    • vi config_wendy_«new».xml change what is not the same
    • Restart the server with the new config
• try to reset the board
  • cd /det/dcs/Development/ATLAS_DCS_LAR/bin/tools/scanSca
    • ./scan.sh -l «link number» -d
    • ./reset_from_elink « address ltdb »

Degraded mode (MTx error) :

Before doing this procédure try to configure the ltdb several times, restart server and retry several times the configuration. If the error on MTx always there then do :

• Go to cd /det/dcs/Development/ATLAS_DCS_LAR/bin/OpcUaLarLtdbServer/mapping/ltdb< num >
• vi configFileLtdb< num >.xml
  • example issue on MTx 4 of SCA 2

<ltdb>
   <sca id ='2'>
      <mtx id='4' mode='degraded'>
      </mtx>
   </sca>
</ltdb>

Calibration file (MDAQ Error) :

You need to check if the calibration file of this ltdb exist on git https://gitlab.cern.ch/agrabas/ltdbdata

• If is existe ask Etienne to add to the set up at P1
• If not :
  • Go to cd /det/dcs/Development/ATLAS_DCS_LAR/bin/OpcUaLarLtdbServer/ltdb_calibration_files/
  • Create a ghost link by making
    • ln -s < file copie > < new file >

Locx2 error :

Go to cd /det/dcs/Development/ATLAS_DCS_LAR/bin/tools/standaloneLtdb Do ./locx2_readRegisters --address opc.tcp://pc-lar-felix-ltdb-XX:48020 --ltdbId X --scaId X

• If you have several Locx2 on the same I2C bus power cycle the crate should be resolve this issue
  • Restart OpcUaSca (SCA reset)
  • After power cycle restart server and configure the board to see if the issue is solve
• if you have only one locx2 is no reachable
  • BNL

_kill server (sca/lar)

• cd tools/standaloneLtdb/
  • ./locx2 readRegisters --ltdbId < num > --help
  • ./locx2 readRegisters --ltdbId < num > --address opc.tcp://localhost:48020
  • ./locx2 readRegisters --ltdbId < num > --address opc.tcp://localhost:48020 --scaId < num >
  • ./loxc2_writeRegister --help
  • ./locx2_writeRegister --ltdbId < num > --address opc.tcp://localhost:48020 --scaId < num > --locx2Id < num-1 > --value < … >
  • ./locx2_reset_

One sca not responding to one elink :

cd /det/dcs/Development/ATLAS_DCS_LAR/bin/tools/scanSca
./scan.sh -l < num > -d

• try to do a power cycle on the felix server and restart all server

Gbtx not well starting