TO DO List for Nerses

Items marked by (D)one, (O)pen, and (P)partial.

2008

D Attempt to change clon01

We have to change the in ~/.cshrc $CLAS/.setup and $CLAS/R3.14.8.2/.setup to exclude clon01 in the if branching.

But, some IOCes are booting with epics account and others with clasboot. Should be standardized and documented. Sergey suggesting to have clasboot for new epics and epics for old one.

Resolved:

By default the new epics environment would be set up.
If one wants to have old epics environment should issue command setup_epics_R3.14.3
No, confusion from IOC standpoint, since their environment is decided by ~epics/<IOC>.cshrc

D LAC HV configuration

Problem: LAC configs were added in the $APP/hvca/db (R3.14.8.2). But ALH for HV does not recognize them.

cd $APP/hvca/db && \
make clean && \
make && \ # will make only *.alhConfig for targets listed in the default target
cd $APP/general/alh && make hv.alhConfig # will create alh config for HV-es.

A PHONY target alh was added into $APP/hvca/db/Makefile which allows to shorten above commands to:

cd $APP/hvca/db && \
make clean && \
make && \ # will make only *.alhConfig for targets listed in the default target
make alh # will create alh config for HV-es.

If this is final we should check-in and update working version. Assumes:

cvs up Makefile # to merge Makefile
cvs add lac1.dat lac2.dat lac.grp
cvs ci  lac1.dat lac2.dat lac.grp Makefile

O Insert clonpc{1,2,3} and camac2 into ioc_alive.adl

This means also to compile HallSystem from $BASEB/src on linux and vxWorks.

In general assumes Makefiles in the $BASEB/src should be modified in order to be able to compile using features present in th R3.14.8.2.

P Driver support for CAEN board 1932

Now, we have working version, which is generic for all known boards. But, because there is one software PRIMARY channel which sets limits for all channels, the interface is kind of fuzzy. Later we probably should separate 1932 board interface. The PRIMARY channel is not configured in $APP/hvca/db/<whatever>.db .So, shift takers will be obscured from internals of this board. But, experts should be aware of internals CAEN A1932. They have to connect using telnet like:

telnet hvpcal 1527

Then enter username and password of mainframe and change PRIMARY channel settings.

• Upload this version into CVS.
• Then upload sy527 version for vxWorks.
• Then upload merged sy1527-sy527 version into CVS.

D Software IOC startup configurations

The software IOC-es are running on:

clonpc1 - gamma and tpe profilers
clonpc2 - HV of beam line devices (hvDvcs CAEN MF)
clonpc3 - test setups for CAEN HV MF-es (hvPCal and hvTest)

Startup directories now are $IOC/`hostname -s`. These files are used:

/usr/local/clas/release/current/R3.14.8.2/EPICS/ioc/clonpc1:
 envPaths
 softioc.pipe
 startup.all
/usr/local/clas/release/current/R3.14.8.2/EPICS/ioc/clonpc2:
 envPaths
 softioc.pipe
 startup.all
/usr/local/clas/release/current/R3.14.8.2/EPICS/ioc/clonpc3:
 envPaths  
 softioc.pipe
 startup.all

Startup script is /etc/init.d/softioc and it is different for each compter. Basically, because software IOC requires active STDIN (otherwise it dies), we make FIFO or named pipe, and using cat we are connecting it to the STDIN of software IOC. Besides, it is nicer than "sleep 1000d", it allows to send commands to software IOC (only once, because EOF or some signal is sent, which means exit to software IOC). This allows to avoid kill process, which leaves active the semaphore. To protect softioc.pipe files they made with read-only permissions.

D classc2 recovery

We suspect something wrong with scalers.

Eugene Pasyuk fixed. One of motor controlling scalers was the cause.

P CAEN EPICS support for EC

DB and MEDM files are generated. Alarms are generated too. Should be checked if the "restore-save HV parameters" features are working.

2010

P Upgrade of EPICS and fixing environment for all machines

Which version should be installed 3.14.8.2 of latest 3.14.11?

• Not much things were changed from 3.14.8.2 to 3.14.11.
  • Array Subroutine Record (aSub) added
  • Added Perl5 CA library
  • Requires GNU make 3.81+, parallel make is now supported
  • Calc expression enhancements (CALC, CALCOUT records)
• MIN,MAX functions support multiple arguments
• Assignment operator :=
• Inf and NaN are valid values
• NOT was changed from unary minus to integer bitwise not
  • New architectures are supported among them linux-arm used by Hovanes for..
  • Support of vxWorks 6.x was changed significantly
  • ai, bi, mbbi and mbbiDirect records support simulation now

Currently, clon02 is upgraded too. On all clon machines by default the new EPICS R3.14.8.2 will run. In order to work with old EPICS environment use setup_epics_R3.13.4 command.

What should be done

• Compile a New version
• Compile HallB EPICS applications
• Test IOC-es with vxWorks v5.x and v6.x
• Test all HallB applications

P MEDM to EDM transition

Compiled, configured, working. Description, instruction notes are here EPICS: EDM compilation and configuration. Make sure:

• Support for semi-automatic logging of any window will not suffer.
• EDM screens from MCC are accessible
• Automatic generation/covertion of adl files to edl ones.

P CAPFAST to VCDT

• Broken Makefile was fixed.
• VDCT is test should be configured for our DB/DBD files.
• Convert capfast files into vdct ones.

Tested, is working with newest java runtime environment. Still to be configured.

O Bug fixes in the CAEN HV driver

Still to be done.

• In the advanced parameters there are some those are not possible to manipulate
• The mainframe status could be different in the EPICS and in hardware if it is not switched on/off from EPICS

D Harp scan analyzing software upgrade

Done. Is in the production.

SLM automatic calibration

Should have the following algorithm:

• If Ibeam(HLB:bta_main_cur) > Ithreshold=0.1 and SLM is off then start calibration
  • Look for a beam trips
  • If a beam trip then
    • Record the SLM vs Ibeam
    • Fit SLM vs Ibeam
    • Obtain the slope and offset and update
• Save the history ?

Beam semi-automatic calibration using 2HO1

Should have the following algorithm:

• Ask MCC to move beam with 0.2 mm steps

and record EC total rates

• Look for central position at 2HO1

A new driver for DVCS RTDs to get rid of LabView

• Develop a new driver for EPICS softIOC for reading out the DVCS RTD values
• Will make the monitoring more reliable

Backup of motor settings

• The backup of values should be automatic.
• The backup should not store the same value if it is not changed.
• The restoration should be manual.

What if burt will read initial values and set them. And experts should just to update latest settings. Will it work?

Foolproof restart buttons on ioc_alive.adl screen

In order to avoid accidental clicking on restart buttons of IOC-es, a foolproof way should be implemented. For example a new window should pop up and ask if operator is sure to restart a particular IOC.