November 7, 2007 online meeting minutes

present: Sergey Boyarinov, Sergey Pozdnyakov, Nerses Gevorgyan, Eugene Pasyuk

We discussed last 2 weeks G9A run preparation activity, following projects were completed:

• online reconstruction code was changed significantly: tagger package, start counter package, TOF, calibration constants; now cmon can perform photon run particle ID on reasonable level of efficiency; we have good target [/Docs/ST_TOF_L1.ps image] from the begining of the run, and Franz will try to implement data monitoring based on two-particle events; more check needed for new TOF implementation job was completed by Franz Klein, Sergey Boyarinov and Eugene Pasyuk

• Nerses and Sergey finished incorporation CAEN HV mainframes (old vme-ceanet-beased and new unix-based) into epics including heartbeat-based monitoring (so-called HallSysterms) and save/restore facility (so-called burt); in ioc status GUI we have clonpc1, clonpc2 and camac2 ioc's controlling those HV mainframes; new LAC HV ioc camac2 works well, no problems reported so far

• Sergey B. switched from 'cdev' to 'ca' epics channel access all related packages in $CLON area, mostly in 'epicsutil' and 'runlog' directories; some problems were reported after several days of running, they were fixed, everything looks reasonable now; the list of epics variables still need to be up-to-dated, Sergey is working on it, will be completed before production starts

• two new monitors were installed, now we have 3 screens on clon02, 3 screens on clon03 and 4 screens on clon04; graphic problem was observed in runcontrol gui, canbe related to the recent changes, currently under investigation

• several days of running shows good online components behaviour, including level2 components; we do not quite understand observed level1 trigger rate yet

We still need to complete following before production starts:

• understand triggering conditions; we'll try to use MOR in ASYNC, or ST&MOR in ASYNC, most likely on Thursday; Eugene will find repaired CAEN discriminators; we still have discripance between epics and daq scalers, but based on daq scalers we have many accidentals; TOF only trigger shows 8kHz without level2 and 800Hz with level2

• cleanup EPICS channel list

• level2 segments requires constant attention, sometimes access to the Hall requires for reboot

• Nerses is considering to use 'medm' from new EPICS in old EPICS

Long-term projects:

• burt-related issues: make burt to save non-HV values (for example electron or photon run), cron jobs to run burt time-to-time (were running in the past for motors ?); Nerses and Sergey B. will look at that

• have one epics_monitor storing data somewhere based on config file(s); run_log_begin etc must get from there, not from epics ca directly; probably epics archiving can be used for that purpose (we'll look at epics archiving anyway, burt or built-in)

• Sergey P. condicting inventarization, storing data in ascii format; in the end all data will be stored in database, we have to deside on interface to that db

• new clonweb and clonmon machines arrived, will install and work on ROOT-carrot and related stuff