Minutes from the 04/13/05 LCLS SLC IOC meeting: (1) Cluster Status Service and Async Utilities: RonM has coded all but two of the functions for the cstrAsync thread and is currently testing. He will have his documentation up-to-date today. RonM will finish cstrAsync and start cstrHdlr when he gets back from vacation next week. Stephanie asked for a change to slcAsyncMeterDbSend. KenU added CSTR defaults for our test SLC IOCs to defaults.dbs and have XL01 to XL05 use it. Ron adjusted CSTR HSTA and CMSK to reflect our current configuration. Tasks left for Ron: * finish cstrAsync and async utilities used only by cstrAsync * cstrHdlr and async utilities used by cstrHdlr (2) New Test IOCs: KenU added XL04 (cdvw4) and XL05 (cdvw5) to the SLC database and SCP panels for RTEMS testing. Debbie and Doug set up the borrowed SPEAR RTEMS CPU as cdvw4 and Debbie is testing. She's having a problem opening a file. Stephanie to investigate. (3) Diagnostics: Diane is adding SLC IOC diagnostics EPICS PVs. She's using special device support (instead of subroutine records) to implement them. She's done event records for stopping and restarting the SLC interface. She's adding ai's for various counters. The plan for these diagnostic records are that they are INST_IO with DTYP=SLC. The INP (or OUT) link will be "@" where the will be all upper-case and 12 chars or less and will indicate the thread and variable name (like "EXECNUMRESTS" for the number of restarts by slcExec). The record names will be IOC::1: which meets the LCLS naming convention. The associated files will be here: slc/libsrc/exec/devSLC.c and devSLC.dbd (dbd installed) slc/saIoc/src/Makefile (change to include devSLC.dbd) slc/saIoc/Db/slcai.db, slcevent.db, slc.template slc/saIoc/Db/Makefile (change to install db's and create/install slcXL0x.template, x=1 to 5) slc/saIoc/edm/ Diane will remove the obsolete subroutine code and database. All startup files will now include: dbLoadTemplate("db/slcXL0x.template") (where x is appropriate for the IOC) instead of the existing dbLoadRecords line. Tasks left for Diane: * support Debbie's RTEMS testing * diagnostics * correct thread priorities (waiting on requirements) (4) Database Service and Utilities: Debbie finished last changes to dbupdate. She's made a few changes to fix bugs found by Ron in his testing including returning proper bad status on certain failures and correcting logic for deallocation of memory in the free list during an SLC shutdown. She's updating her documentation. She's changed her code to use the proper message codes. We discussed adding new message codes but decided, in the spirit of limiting our scope of work, to use existing message codes and just find one close enough to the condition that needs logging and use the proper text. Debbie discovered that the SLC IOC will wait forever for the SLC database to be downloaded. There is no watchdog, like the SLC micro, that will implement a timeout. On the SCP side, eventually the "reboot" times out and the IOC is set offline. Also, PARANOIA will set the IOC offline when it doesn't get replies to its requests. Right now, to get out of the wait, the whole IOC needs to be rebooted. We discussed changing slcExec to respond to stop/restart messages while it is waiting for the database threads to finish its download. We think we'll add this function later if it becomes an issue. Tasks left for Debbie in this order: * test on RTEMS * add any missing diagnostics (collision count, for one) * update documentation (5) Magnet Service: Stephanie provided a list of which side (EPICS or SLC) is the master and thus has control of a setpoint under various operating conditions. (6) BPM Service: Yesterday, Mike Zelazny showed Diane and Debbie the SCP BPM acq user interface. This is the first step in gathering requirements.