Minutes from the 07/26/05, 08/02/05 LCLS SLC IOC meetings (updated for Aug 4 progress): (1) Updated Task Lists: Done Tasks since 7/12/05 Who When ------------------------------------------ --------------- ---------- * SLC test databases Nancy Jul 26 * Decide on proxy config for Oct test Terri Jul 26 * Prototype mgntHdlr in CVS Kristi Aug 2 * More slc and db utilities for Hdlrs Kristi Aug 2 * Change dir struct to make EPICS side Steph Aug 2 more independent of SLC * BPM Test DGRP MikeZ Aug 2 * bsacHdlr - msg receive/reply, Debbie Aug 3 dummy calibration to make SCP happy * IOC EPICS bsac test databases Steph, etc Aug 3 * SCP update - ignore checksum MikeZ Aug 4 Major Tasks Who When ------------------------------------------ --------------- ---------- * BSAC code/test finished Debbie, Diane Aug 17 * Magnet SLC design review Kristi Aug 30 * Magnet EPICS design review Kristi Aug 30 * Change iocLogServer for CMLOG Steph Sept * Decide on SLC PNET timing diags support Steph Sept * Decide on SCP wire-scanner Controls Group Sept * ICD - EVR (w/Dayle) D, D, Kristi Sept * ICD - BPM (w/Till) Debbie, Diane Sept * ICD - Toroid and GADCs (w/Doug) Debbie, Diane Sept * ICD - Fast Feedback (w/Steph) D, D, Kristi Sept * ESD - IOC naming and restrictions Steph Sept * Magnet code/test finished Kristi Sept * LCLS SLC modelling database setup, etc Mark Woodley Next year Loose End Tasks Who When ------------------------------------------ -------------- ----------- * BSAC test EDM displays and simulator Steph Aug 9 * Get rid of slcExec and cstrHdlr kluges Steph, Diane Aug 12 * slcCmlog change to honor stop flag Steph Aug 12 * Determine how BPM ANGL,OFFS,XTOL used Diane, Debbie Aug 17 * Thread priorities Steph Aug 30 * Add how-to-add-an-slc-aware-ioc Steph Sept * Timing of 10 sec update not right Steph Sept * CSTR/ASYNC code review Steph Sept * Wrong IPL time when not booted from SCP Steph Sept * slcCmlog error metering Steph Sept * Pulse ID in SCP correlation plots MikeZ Oct? * Update specs All As needed * Update record naming conventions All As needed * Update database standards 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) (2) Vacation: Diane - July 25 to Aug 5 BobF - Aug 5 to Aug 15 Kristi - Sept or later Steph - Aug 15 to 23 Tony - gone until Oct MikeZ - Aug 15 to 24 (3) BSAC Progress: Debbie is parsing the incoming BPM meas prep message and filling in EPICS meas defn records. She is responding to calibration message with dummy data to make the SCP happy. Debbie's goals for next Tues: * finish dummy calibration work * finish meas prep and MDEF record updates * finish collection area preparation Debbie will add the SCP measdef ID to an EPICS MDEF record for display purposes. It was suggested that measdef information that is stored in EPICS MDEF records also not be stored locally in memory - why put the same information in 2 places. Diane to look at the reply message when she gets back from vacation. BTW, Diane will have XL03 has her own test IOC since Bob Sass is no longer using it (he's got the new LI36 instead). She can run mod_gs if she doesn't like the node name it's assigned (currently, noric06). When a unit doesn't exist in the EPICS database, a message is logged and bad status returned. The bsac task will not exit on this condition. Considering adding a 1 second BPM async check function for bsac to use for HSTA consistency checking and possibly the measurement timeout check (instead of an internal timer). Mike mentioned that BPM requests may be more than the 64K maximum and that Tony added special code to handle >64K. We will not support >64K. We don't think we need that capability. Currently, we do not keep our test IOCs up and running. So Mike cannot test when we are not around to start an IOC for him. Mike needs instructions on how to start one of our soft IOC. Currently, when selecting the BPM LCLS test DGRP from the SCP, the SCP delays as it tries to talk to every micro that it thinks is online as a result of dbgen checkout. Is there a way to disable this "feature"? Currently, the existing SCP pulse ID "acquisition" software does not work. This uses a special primary called "PHYS" and a special secondary called "PLSD". Does MikeZ need to get this running by the Nov FFTB tests? Patrick mentioned he wanted the ability to plot or dump out pulse IDs from the SCP correlation plot facility. (4) BSAC Databases and Displays: Nancy has the test SLC databases ready. She added 2 ARRY units which gives us 24 gated-ADC channels, 6 BPMs, and 2 TOROs. Steph has the equivalent test EPICS databases ready. Steph's goals for next Tues: * finish EPICS test database checkout and simulator (Greg White suggests a noisy dirty sine wave for BPM data) * EDM test displays Note the simulator will run at 10hz. MikeZ currently has no time to help with any BSAC database or display work. Mike created a BPM DGRP for the test IOC BPMs. When testing BPM buffered acquisition, one test would be to change the DGRP to have different order of unit numbers. The BSAC databases and subroutines will reside under epics/site/src/bsac instead of epics/ioc/bpmApp where we expect Till to commit his work. This will provide better separation between bsac work and Till's work. In setting up the test databases, Steph has allocated 6 special meas defs for: 1Hz 10Hz Full-rate Single-shot Fast Feedback 1 Fast Feedback 2 We don't anticipate more than 2 fast feedback loops per meas IOC (usually zero or 1). (5) Magnet Progress: Kristi committed a bare-bones prototype mgntHdlr and other mgnt files. Testing is needed. She also committed some utilities. She is also starting some prototype magnet databases with some soft interlock and on/off records. All her EPICS database and software will reside under epics/ioc/mgntApp. Kristi's goals for next week: * mgntHdlr testing - msg receive and reply * more mgnt database development and diag EDM displays Nancy has the test SLC database ready for magnets. She provided LGPS, QUAD, XCOR, YCOR, and BNDS. As decided in the reqts review, there will be no special kicker support in the SCP. What kicker information is needed for Mark Woodley's SCP modelling? Nancy wonders if we'll need to control any amplitudes (AMPL) or phases (PHAS) from the SCP (ie, for LLRF). We think we can support them if needed. Steph added files and changed configure/RELEASE so that each IOC app that needs to be SLC-aware only needs to add a few extra lines to their Makefiles. (6) Higher Level Message Utilities: Both Kristi and Debbie say their code would be simpler is there was a higher level layer on the message utility. Steph sent out a proposal. Wait for Diane to review. (7) Record Naming and Database Standards The location of the device, not the IOC name, is in the record name. Bob Fuller thinks this is OK. Need to confirm with Mario. This affects documentation and labelling. Since database templates are loaded per IOC, templates will contain the IOC name. The VDCT standards need to be updated. We also agreed to remove the recommendation that all records in a database fit on one VDCT screen. Two screens seems to be more workable.