Monitoring Controls System meeting on February 18, 2003

Attendance:
Steve Hahn, Andy Hocker, Mark Knapp, Yeongdae Shon, Irina Shreyber, Margherita Vittone, Marcin Wolter, John Yoh, JC Yun

NEWS:
Since the 3 week shutdown, the instantaneous luminosity at B0 hovers around 50% of those of pre-shutdown. There have been frequent quenches recently. We expect 2 day access to fix COT high voltage problems in a week or so.

SYSTEM:
iFIX upgrade went pretty well in overall. We still have 3 computers to be upgraded during next available downtime. These are SVXIICON, CDFPEHV and CLC1. 

VOLTMAN ... OK 
CDFTEMP ... OK 
COT1    ... OK(?)
PISABOX ... OK ... needed Visual Studio 6.0
MUON3   ... OK ... Visual Studio 5.0 worked with linker 6.0 
               ... 10-20% slower now

CDF51683 ... Not OK
         ... not using compiled excutable of VB and x10 slower
         ... had to go back to V2.21
         ... Oleg is working on building a compiled excutable

SERVERS ... Cryo area servers working better after all servers upgraded
Steve added that the most of the tags in iFix database need not to be tags. Since there are so much overhead for each tag, we might better off by making data files to write information. This is too late for the most of the systems, however.
  • Steve's transparacies.

    • Mark pointed out that it's too early to gauge the system performance before all the computers are upgraded.

    DOCUMENTATION:
    Contents of Jonh's transpanacy : 

           Detector subsystem people

        J. C. Yun       --iFix
        Steve Hahn      --ACNET, CEM, CHA, WHA, GAM
        Bob Wagner      --COT
        Phil Schlabach  --Muons
        Karen Byrum     --CES, CCR, CPR
        Jonathan Lewis  --Trigger Inhibit
        Andy Hocker     --SVX
        Oleg Poukhov    --Muon Scint
        Gerry Bauer     --TOF   
        Willis Sakamoto --Plug Cal
        Margherita Vittone --Icicle
        Andrea Bocci    --MNP, BSC, RPS
        Jaco Konigsberg --CLC
        David Dagenhart --PTM
        Marcin Wolter   --PSM
        Mark Knapp      --PC Backup

Now that the access is finally over (we hope), let's 
make sure that the MonDoc (Monitoring Documentation web pages-- 
linked off the monitoring Ace page--as iFix, detectors) are Up-To-Date 
so that the Shift people will have all the proper instructions and
recovery procedures for each system readily available. 

Many of the detector subsystems have web pages
that are more than six months old, and thus could be outdated. 
Please update your web pages to reflect the current status, especially 
the "Recovery procedure web pages"--please put the current date on those 
web-pages after you're done (or verified that they'sre up-to-date). 

Unless I receive email stating that you have updated your web pages
within the next 10 days, I will be coming personally to ask you to verify
that your web pages are current and up-to-date.
    For those who do not know how to get to the CDF Hardware Monitoring Documentation page, use the following url:
    http://www-cdfonline.fnal.gov/mcs/mondoc.html
    We desperately need instructions to shift crews for the Trigger Inhibit. Also we would like to beef up ICICLE instructions to include recovery procedures from some of the specific error states.

    TRIGGER INHIBIT:
    There have been several incidents when trigger inhibit stopped the run control even if there was no obvious alarm. One plausible explanation by Steve is that the Camac logic module which process all the hardware trigger inhibit signals are flaky. On those cases all the systems set off trigger inhibit for a fraction of seconds. Since we do not have spare module there is no immidiate solution. Jonathan et al. plan to install a VME system for this in the future.

    PLUG:
    Margherita said Luciana and herself tried trigger inhibit test but could not succeed. They will try it again in the near future.

    SVXIICON:
    Andy is waiting for CDF downtime for the iFix upgrade. It will happen most likely during the access for the COT fix. Since there are several systmes (CLC1, Solenoid2, SVXIICON) using smart socket it would be useful to have a common area for the utilities needed for smart socket. We need to figure out how to transfer the code to the common area.

    VOLTMAN:

    Marcin said the 10 second readout cycle could be reduced to 6 seconds only if we buy very fast CPU. Also he noticed the pictures opened at the Vnodes are vertically elongated. JC volunteered to contact Jim Lostot for this problem. Marcin said he could not make the alarm list working since the software was upgraded. Andy said it never worked for SVX system. We need to look into this since it's very useful for diagnostic purpose.