Minutes from the 10/13/04 LCLS SLC IOC meetings.

Next Meeting: 10/27/04, 9:30am, in B5, second floor.

(1) Data File with Database Information - Ken will be generating files
    with additional database information to supplement the downloaded database.
    This will be done during the DBGEN and the files copied to production
    during the DBINSTALL.  The first file will map the primary names with 
    numbers.  The second file will map each primary/secondary with 
    secondary number, datum format type <A,I,R,S,T,Z>, datum size in bytes,
    and number of data items (V == variable).  Debbie will read in the
    files as part of the database download and keep information in memory
    with appropriate hash tables.  Debbie will develop routines to access
    the information in these tables.

(2) Place for Data Files on NFS Server - Debbie will send requirements to
    Ken Brobeck for a place in NFS for the extra data files.  We will need
    a development area and production area.

(3) Database Access API Review - the dbio prototypes were reviewed.
    Debbie says it'll take 2-3 weeks to finish the dbio routines.  A short
    list of routines are included that can be called at the iocsh prompt
    including a "dbedit" routine that is similar in functionality to dbedit
    on VMS.  This will allow us to change ST1 data without requiring an IPL.
    dbgettype, dbdump, dbedit are highest priority of the iocsh routines.
    http://www.slac.stanford.edu/comp/unix/package/epics/lcls/slc_ioc/dbio.txt

(4) DBEDIT Discussion - should we make who can change ST1 ("static") data in 
    the IOC SLC database be configurable?  A thought for later - DBEDIT on VMS 
    could be expanded functionally to write files that the IOC could use to 
    update its database without requiring an IPL.

(5) "Micro IPL" and "Micro Reset" - Tony has added new messages to the SCP
    that the IOC will use to do an SLC task restart/database download and SLC
    task stop.  He has also added a messages for DBEX starting and exiting.
    The DBEX start message also includes the ASCII database's current version 
    number.  Since we are restarting threads/tasks instead of the IOC,
    Debbie and Diane are careful with resource allocation and freeing.  
    Diane will study EPICS utilities to check that resources can be freed on 
    thread/task restart.  Diane will look into using the EPICS database and
    a subroutine record to do the same functionality from CA.

(6) Locking of Data in the SLC Database - Since ST2 (setpoint) data can be 
    changed by more than one task, Debbie needs to lock of the data while the
    updates are done.  How to lock - how much of the database should be 
    in one lock set?  We may need locks for ST3 (readback) data - it depends 
    on the design of the tasks that update ST3.

(7) New error function code "IOC_CODE" - RonM is adding a new function code
    to cmlog fwdBro and VMS FMT_MSG to allow forwarding of text messages to 
    SCPs.
 
(8) Time Conversion Routines - Diane is done (net_get_time, difft10 (msec), 
    and difft10u (usec).  We see no need to have a VMS time to EPICS time
    conversion.

(9) Superblock 0 Conversion - Debbie is done.

(10) Rebooting via PNET - the PNET broadcast include micro-reboot flags.
     Do we need the ability to reboot SLC IOCs (that have PNET modules) using
     this broadcast?