SLC-Aware IOC Meeting Minutes - 10/04/05 plus 10/06/05 notes: (1) BSAC Status: * Diane reports a problem with the bsacAbort logic on Linux that causes a crash. For now, she has added a workaround so the crash doesn't happen. * There is no xupport for multiple concurrent BPM acquisitions for a single SCP. * There is no support for nice error messages when there is not enough room to send back all the data in an acquisition (ie, when nrpos (MEASCNTMAX) is big and the IOC has many ARRYs). The SCP reports some non-intuitive messages. Nothing crashes in this case though. * BSAC large message replies done and cleanup in-progress. No significant BSAC work planned until integration testing. * There is a problem with downloading the SLC database while PNET 360hz interrupts are happening. Kristi volunteered to take a look at the PNET driver while Dayle is away. Kristi and Steph to temporarily change Dayle's test PNET IOC (in her office) to use Dayle's evrTest so Kristi can debug here instead of B5. (2) SLC-Aware Magnet Status: * Function priorities: (1) TRIM and CHCK for non-strings first (2) TRIM and CHCK for strings second (n) Calibration and calibration diagnostics (from SCP) last or never * Kristi has put all her untested SLC-aware magnet code into CVS. Debbie will review, change, and test after the design review. * Kristi, Debbie, Diane met during the last week of Sept to sketch out the SLC magnet side threads and data flow. It was decided that two threads and queues are needed: (1) mgntHdlr - waits for Alpha messages, checks for an update to HSTA/BDES in the SLC database and writes new values to EPICS, starts an action, waits for the action to complete, replies to the Alpha, and then waits for the next Alpha request. (2) mgntAsync - waits on a queue that is written to by monitor callbacks that indicate either action completion or change to HSTA/BDES, reads the EPICS database and updates the SLC database, and releases the mgntHdlr. The queue will also get the ASYNC function requests. Note - cstrAsync will need to change to send to mgntAsync instead of mgntHdlr. Note - slcExec will need to change to send a STOP request to both mgntHdlr and mgntAsync queues. Note - a new mgnt ASYNC function might be added to check for HSTA/BDES change in the SLC database without a corresponding message that forces update of the EPICS database. In this case, the SLC change will be overwritten with the EPICS values after a certain period. (3) SLC/EPICS Special Interface Records for Magnet/PS: * Kristi plans to add a set of mbbo's and bo's to control magnet control logic action (all types of power supplies will have the same set of records even if not all are used for each type). Logic action will be implemented either via groups of records or sequence(s). She plans to add additional records to go from a HSTA control record (mbboDirect) to these controlling mbbo/bo's and from the controlling mbbo/bo's to a HSTA readback record (a second mbboDirect). This will make it easier for the SLC-side to both change the control records when the SLC HSTA changes and the change the SLC HSTA when the control records change via CA user. * In a similar vein, Kristi plans to add a set of mbbo statuses of various control actions (like standardization) that are updated after the action is done either via sequence or groups of records. She plans to add additional records to go from these mbbo statuses to a STAT readback record (mbboDirect). The SLC side will then be able to read the STAT record at the appropriate times. * The SLC magnet job has a special algorithm to decide if the *ACT (readback) values need to be updated on the Alpha during the normal 10 second periodic update. The algorithm is different from the MDEL logic in EPICS and uses the TOLS field in the SLC database. This checking will be done only on the SLC side and no attempt will be made to keep SLC TOLS values and the corresponding EPICS values (MDEL, warning and alarm limits) in-sync. * The only SLC magnet secondaries that will be updated by mgntAsync: (1) *ACT readback values (* = B,V,A depending on PS type) (2) BMON - second ADC readback when available (3) STAT - status (4) CTIM, TTIM, KTIM, ZTIM - timestamps when specific actions happen (calibration, trim, standardization, lost standardization) (5) ITRY - number of times a trim is attempted (6) IMAX - maximum current for standardization (7) RIPL - ripple readback (if available) (8) *DES setpoint values (* = B,V,A depending on PS type) (9) HSTA - control bits Except for IMAX and RIPL, no I or counts values will be updated in the SLC database. Polynomial coefficients, settling times, tolerances, and other constants will also NOT be kept up-to-date in the SLC database. (4) Status of Prototype PS Controls Appl: Kristi plans to bring up the bare-bones SPEAR etherPSC app with just name changes (no additional records yet) by Friday, Oct 7. This includes the SPEAR EDM diagnostic displays. (5) Updated Task Lists: Done Tasks Who When ------------------------------------------ --------------- ---------- * dbSend infinite loop bug when DBEX sick Debbie Sep 30 * Finish large msg replies for BSAC Debbie,Diane Sep 30 * Change dblist header for FECC Debbie Sep 29 * slcExec to check CSTR CDSC bit for FECC Diane Oct 5 * SCP panels for dev magnet testing Kristi+CPE Oct 5 Major Tasks Who When ------------------------------------------ --------------- ---------- * Prototype EPICS PSC slow controls Kristi Oct 7 * Magnet SLC design Debbie Oct 14 * Production SLC IOC test with PX00 Steph, KenB Oct 14 * Magnet EPICS/SLC design review Kristi/Debbie Oct 18 * SLC IOC test with PNET D,D,S,Dayle Oct 21 * Magnet SLC code Debbie Oct 28 * Continue magnet/PSC EPICS Kristi Oct 28 * Fast correctors - EPICS and SLC K,D,S Nov * Change iocLogServer for CMLOG Steph Oct * ESD - IOC naming and restrictions Steph Oct * ICD - EVR,BPM,GADCs,FBCK All Oct * BSAC integration test with EVR D,D,S,Dayle Oct? * BSAC integration test with ADC D,D,S,Till Oct? * SLC database for injector Kristi,Nancy Nov * LCLS SLC modelling database setup, etc Mark Woodley Next year Loose End Tasks Who When ------------------------------------------ -------------- ----------- * iocsh dbupdate ALL* bug Debbie Oct 14 * Problem with DB download & 360hz intr's Debbie,Diane Oct 14 * Update VDCT/database standards Steph Oct * Thread priorities Steph Oct * Change "BSAC" back to "BPM" on SCP Steph Oct * Timing of 10 sec update not right Steph Oct * LCLS channel archiver Sergei Oct * LCLS channel watcher, setpoint restore Kristi Oct * slcCmlog change to honor stop flag Steph Oct * Add how-to-add-an-slc-aware-ioc Steph Oct * slcCmlog error metering Steph Oct * Pulse ID in SCP correlation plots MikeZ Oct? * Check if regional beam codes affects BSAC Steph Oct * Update specs All As needed * Update record naming conventions All As needed TBD: * Increase max number of micros * Second production proxy? * Production "domain" split * MCC CA server change for controlling CAMAC power supplies (need estimate)