Instructions for running the DAQ
Responsible for this sheet: Vladimir
Rybnikov. Last Update: January
30th 2003. Further information:
DAQ
WWW page
Some internals of the DAQ system can be found in
HERA-B
Run and Slow Control Tour by Vladimir
Rybnikov
What to do if the expert is not reachable? Contact
your favorite DAQ expert: see
expert shift list.
Contents
How to start a DAQ
run setup.
DAQ should be run
from the DAQ terminal (hb-cr10)
closest to the Shift crew terminals.
|
This chapter contains instructions to set up all processes required
to take data with the complete hera-b detector or parts of it. In this
documentation processes are printed bold. The code for buttons
on graphical user interfaces is this.
Names of states are printed such.
prcmanag is responsible to setup the processes
which are common for all ongoing runs. There should be always a main
prcmanag
running
in hbdaq0 .
This process is launched automatically when crcui is started.
While running the number of prcmanag should increase depending on
the run type. The abnormal termination of one of them implies a crash in
one of its child processes (see trouble shouting)
. In this
case a list of killing process message will appear in
the error logger window.
The user interface is called crcui and can be found in the usual
online path. Steps to create
a run
THESE ARE THE INSTRUCTIONS
TO BOOT ALL PROCESSES IN THE RUN.
IN CASE THE SYSTEM IS ALREADY
RUNNING PLEASE FOLLOW INSTRUCTIONS IN change
RUN.
|
-
logon to hbshift on hb-cr10 (passwd is noted in paper logbook).
-
start crcui from
the Button bar at the upper left corner of the display. (RUN_control).
-
press new run on the
popped
up window
-
select the run configuration "switch"
.
-
after selecting the detector and run components ( red
button means selected) press the key OK
and
confirm the pop-up.
-
Kill the yellow Error-Logger Window
with
(Ctrl-C) in case it is there.
-
select the run type in
the RUN cfg menu. Every
option stands for a different Second Level Trigger code. The configuration
of the different running condictions are shown below.
-
FOR PRETRIGGER RUNS --->
Select FLT (should turn red)
in the Run conf menu on the control
panel.
-
press MENU on the Process
Control section of the control panel (a pop-up
will appear).
-
press Create on the pop-up
to create processes.
-
wait till "state:" on the control panel
is INITIALIZED (and watch the two error
loggers; the white one is global and it is runnig permanetly in the
hb-cr10 DAQ node. the yellow one is
specific to the run).
-
the "Trigger Monitor"
will appear to monitor and control rates for non-FLT
runs.
-
A green window"DAQ
Monitor" should appear to monitor the data stream just
before logging.
-
Press PAUSEin the control
panel and you have a new run. The state should change to
STANDBY.
-
Please be patient, the transition could last for few minutes.
-
Enable FLT triggers for PHYSICS
RUNS. (Ignore this point in other cases and execute the next step).
-
Go to RUN pressing
START
button. You are taking data now!!!!!.
-
From now on you can handle new runs following instructions
in change RUN.
-
The FED error monitor provides ONLINE
the state of the subdetector readout. It will generate alarms in case of
severe errors.
-
Normaly there is a predefined
value for the trigger rate..... go to "Trigger
Generator" and type the one you wish.
RUN TYPE CONFIGURATION.
RUN MODE |
TYPE |
FLT BIT |
FLT |
RUN type |
Random Trigger |
Test Trigger |
PRETRIGGER |
PHYSICS |
ON |
ON |
/Production/Physics/Archive/F |
ON(4Hz) |
OFF |
MINIMUM BIAS |
BX |
OFF |
OFF |
/Production/BX/Archive/ |
ON |
ON |
CALIBRATION |
CALIBRATION |
OFF |
OFF |
/Production/Calibration/Archive/ |
ON |
ON/OFF |
INTERACTION |
INTERACTION |
OFF |
OFF |
/Production/Interaction/Archive/ |
ON
|
ON
|
Three logging configurations are available (how
to described here):
LOGGING MODE |
Logging path |
Shared Memory partition |
ROBOT copy |
Comments |
ARCHIVE |
/hb/data/REAL/ |
daq |
YES |
PHYSICS RUNS |
NONE |
NONE |
daq |
NO |
DEBUGGING MODE |
DISK |
/hb/data/TEST/ |
daq |
NO |
TEMPORARY DATA STORAGE |
Change
RUN number:Remember that the run configuration will remain the same.
This is needed in the typical cases:
Changes in any parameter external to the DAQ...(i.e. Target
rate).
A detector in the run stops for a while (change gas,....)
HOW TO CHANGE A RUN
-
Go to the control pannel running in DAQ
terminal (hb-cr10).
-
Go to READYpressing STOP
button.
-
Disable FLT triggers if running.(Ignore
this point in other cases and execute the next step).
-
Go back to STANDBY pressing
PAUSE
button.
-
Enable FLT triggers for PHYSICS
RUNS.
-
Go back to RUN pressing
START
button.
Steps to terminate run processes:
(KILL ALL PROCESSES)
-
This is needed only in case of error
or change of the RUN configuration.
-
disable the trigger in the "Trigger
Generator".
-
disable the FLT trigger (look
in FLT documentation ).
-
Write down the number of events and run remarks in the RUN
checklist.
-
VERY IMPORTANT = Change
state to READY pressing the buttonSTOP
in the run control. Wait until the state in
the run control changes to READY.
-
press MENU on the control
panel and a pop-up will appear.
-
press Terminate on the
pop-up.
Further
Documentation on Standalone running Top
How to Select a run configuration.
Select the RUN configuration.
-
Select switchrun
configuration by clicking with the mouse.
-
Select COMPONENTS in the RUN .
Red button means the
component is selected. (the default includes all components ).
-
FARM enables the logging. It
is mandatory for Interaction and Physics
runs.
-
SLT is mandatory
for
any switch run.
SLT process type selects the SLT code to be booted at the SLT farm.
That means it does not apply before the run processes
are killled and restarted.
Physics -> triggered
events.
Callibration ->ECAL
calibration run.
Interaction ->Event Assembly
with no trigger.
BX -> not enabled.
TEST -> not
enabled.
Logging:
Archive -> data
saved in /hb/data/REAL/... copied by robot to tape.
None -> data not saved
in disk.
Disk ->data
saved in /hb/data/TEST/.... not copied by the robot to tape.
BX -> not enabled.
TEST -> not
enabled.
The different Trigger options are:
-
FLT-> FLT
triggered events.
-
SLT -> disabled.
-
TLT ->disabled.
-
4LT -> disabled.
Further
Documentation Top
How to Start/Pause/Stop
an already setup run
The HERA-B Run Control Panel (crcui) is the top level
run control for the moment being. It contains following important user
controls:
-
Start button to start
or resume a run. If the current state is READY
a new run is started. If the current state is STANDBY
the run is resumed; i.e. the trigger is re-enabled
-
Stop button to stop a
run. If the current state is smaller then READY
the state machine is cycled into READY
-
Pause button to pause
a run. If the current state is RUN
the triggers are disabled and the FCS is reset - but the current run is
not terminated. If the current state is smaller then STANDBY
a new run is started and brought into pused state.
-
Process Control button
to get a pop-up which contains the controls:
-
create button to setup
all processes required to run
-
terminate button to stop
all processes required to run
-
terminate and quit button
to stop all processes required to run and exit panel
-
Dismiss button to exit
the pop-up.
the create and terminate run buttons allow the setup
and removal of processes required to run. They should only be operated
when a data taking poeriod is starting or after the likely event of a system
crash. In this case start from "cleanall switch" in the run
setup instructions.
How to select a RUN type.
To select the RUN type go to Run Cfg in the pannel and
select the Type menu,
SLT process type selects the SLT code to be booted at the SLT farm.
That means it does not apply before the run processes
are killled and restarted.
Physics -> triggered
events.
Callibration ->ECAL
calibration run.
Interaction ->Event Assembly
with no trigger.
BX -> not enabled.
TEST -> not
enabled.
Logging:
Archive -> data
saved in /hb/data/REAL/... copied by robot to tape.
None -> data not saved
in disk.
Disk ->data
saved in /hb/data/TEST/.... not copied by the robot to tape.
BX -> not enabled.
TEST -> not
enabled.
The different Trigger options are:
-
FLT-> FLT
triggered events.
-
SLT -> disabled.
-
TLT ->disabled.
-
4LT -> disabled.
Further
Documentation Top
Trigger Generator Control Panel
The Trigger Generator Control Panel allows to control
the random trigger and displays the data taking rate. The Controls are
only needed for non standardt interventions.
The controls of the Trigger Generator Control Panel
are:
-
EVC Monitor launch the
trigger display
-
Parameters launch the
Trigger Parameter Control window with the following fields. Clicking the
button enables the trigger....then the field should turn red.
-
RATE PROTECTION fixes
the maximum event rate handled by the Event Controller . It is the maximum
rate
the Second Level Processors will accept.
-
TEST TRIGGER RATEcontrols
the rate of an external pulse generator.... it sees the secopnd level trigger.
-
RANDOM TRIGGER RATE controls
the rate of the random trigger. It does not see the second level trigger....
it provides a sample of minimum bias in any condition.
-
Close to terminate the
expert window
-
Exit button to leave
it (shouild not be used)
-
Rate field displays the
last measure rate displays in the SLT input Rate
.
The two monitors show the status of the DAQ rate:
-
"Trigger Rate Plot" shows the input rate to the second
level farm as a function of time The Rate should be rather smooth; if the
rate gets unstable decrease the atempted rate in the "Rate" control field.
-
"EVC Dead Time Corrected Rate"
shows the input rate to the second level farm as a function of time corrected
by the Event Control dead time.
-
"Number events" shows the total number of events since
the panel has been started.
Fast Control System counters monitor.
The FCS monitor displays the rates of events that are rejected
for different reasons . At the same time it is an easy way of displaying
the actual settings of the
FCS mother and daughters. This monitor is normally runnning
in the upper terminal attached to hb-cr10.
TRIGGER
MASK
The definition
of the trigger masks as they are generated at the Fast Control System:
TRIGGER MASK |
HARDWARE MEANING |
COMMENTS |
0 0000 0001 |
LEMO 1 |
ECAL Internal LED calibration |
0 0000 0010 |
LEMO 2 |
Front Panel trigger for test |
0 0000 0100 |
LEMO 3 |
Front Panel trigger for test |
0 0000 1000 |
LEMO 4 |
Front Panel trigger for test. |
0 00010000 |
VME |
Standalone trigger |
0 0010 0000 |
RANDOM |
Random trigger |
0 0100 0000 |
FLT |
FLT trigger |
0 1000 0000 |
PRNDM |
Pseudorandom trigger |
1 0000 0000 |
TP |
Test Pulse trigger |
The example of composite trigger mask:
Trigger mask
|
HARDWARE MEANING
|
0 0010 0001
|
LEMO 1 + RANDOM
|
0 0010 0010
|
LEMO 2 + RANDOM
|
0 0010 0100
|
LEMO 3 + RANDOM
|
0 0010 1000
|
LEMO 4 + RANDOM
|
0 0000 1111
|
LEMO1+LEMO2+LEMO3+LEMO4
|
FED Error Monitor
DAQ Readout Monitor
The DAQ monitor looks at the event structure and reports
information about the event content and errors. The display has three sections:
-
The upper one displays information about the run configuration,
run number, experimentnumber,....
-
The most-left reports on global errors and displays
event information.
The most-right reports errors and displays information
per
component. In this two columns numbers in reverse video show
total
error counts for the ongoing run. Normal printing denotes errors
per event.
How to detect problems:
-
error counts of zero indicate a perfectly working system.
-
error counts in reverse video do not indicate a fatal problem
but rather an almost well behaved system.
-
ECAL/OTR/MUON has faulty FED's. That turns into few reported
errors per event. This is considered to be normal.
-
SVD has spuriously faulty feds. Therefore you see the total
error count (in reverse video). This is OK.
-
If the error count per event increases or becomes comparable
to the number of feds (column #FEDs) there is a problem. In this case a
new run should be going to READy and back to RUN using the run
control.
How to know
that a run setup is there and ready
Top
NAME SERVER
This tool displays the process names
running
under the HERA-B DAQ enviorement. (Be extremely
careful when you use it, since you can disturb other people work).
How
to get online errors reported.
The error logger "erwin" is started by typeing "erwin"
at the command line. It displays a window which contains the last some
error/warning print outs. In general an error entry looks like:
17:28:58 [Error]
evc
- Flt-id hasn't changed after 3 retries
-
The first field is the time the error
occured at
-
The secoond field is the severity level
-
Inform for run progress
information
-
Warning for signs of
potential misbehaviour which do not affect the dta taking from a DAQ point
of view
-
Error running is seriously
effected (data most likely useless)
-
Fatal continue running
is impossible
-
next the process produceing the error
message is shown
-
then a message string describing the
cause of the message
Further
Documentation
Top
Trouble shooting
Action to be taken when some of the DAQ
processes die.
During the running few
processes are allowed to die. A pop-up yellow window will tell you that
the process died and offers you three options:
-
Ignore, the process will be ignore from there on, the process is not critical
for data taking.
-
Restart process, the process will be restarted and will run, the process
is not critical for data taking but recomendable (monitoring).
-
Kill the RUN. It is needed when the missing process could affect
the quality of the data.
Known
actions to some dead processes:
Process Name |
Action |
/RUN_switch/SLP/_* |
Ignore |
/RUN_switch/ECAL/PC_0 |
Ignore |
/RUN_switch/SLT/PC_0 |
Ignore |
/RUN_switch/SVD/PC_0 |
Ignore |
/RUN_switch/SVD/dmon_slt_svd |
Restart |
/RUN_switch/*/dmon_* |
Ignore |
/RUN_switch/ECAL/FLT/pt_moni.hb-vme* |
Terminate RUN.( Contact ECAL expert) |
/RUN_switch/ECAL/FLT/pt_init.hb-vme* |
Terminate RUN.( Contact ECAL expert) |
/RUN_switch/ECAL/fed_init.hb-vme* |
Terminate RUN.( Contact ECAL expert) |
/RUN_switch/ECAL/status_agent |
Ignore |
/RUN_switch/RHP/gatherer_4lt_* |
Ignore |
/RUN_switch/FARM/l4aSender_* |
Ignore |
How to cure few known problems.
-
If after stopping a run you don't get the green window telling
that you can start a new run:
-
Wait a bit!,... it could take up to a couple of minutes.
-
Remove from the name server the
process /RUN_switch/runcleanup.
-
If the system is not logging data (no
data appears in the Daqmon window ).
-
First test that you are not running the ECAL
calibration.... it does not log data.
-
Check the FCS rejected rate and
SLT
input rate.
-
If the SLT trigger input rate is non zero and FCS rejected
is too high (> 50 KHz), most probably is pretriggers/FLT, check MUON/ECAL
pretrigger and FLT documentation. In this case the SLT rejects all the
events and nothing moves to the logging.
-
If the SLT triger input rate is zero...try to stop/start
a run. If it does not work, try to terminate/create a new run.
-
Only one process has been started
and the booting stopped.
-
RUN_WATCH process has
problems in getting all nodes available for
the run.
-
Check up in the /SYSTEM/
error window the node that the RUN_WATCH
has problem with.
-
Try to reset the node
with the help of the Central Crate Control
-
The system cannot be started due to "START-UP
ERROR"
-
The processes started by the /SYSTEM/prcmanag in the
previous run could be still running. The list of the process with the corresponding
nodes in the table bellow:
PROCESS SEEN BY PROSERV
|
NODE
|
DESCRIPTION
|
/RUN_switch/run_watch |
hbdaq1
|
process checking availability of all the nodes |
/RUN_switch/PIDxxx |
hbdaq2
|
main name server /RUN_switch/, (xxx - any number) |
/RUN_switch/rscmanag |
hbdaq1
|
resource manager for the run |
/RUN_switch/RC/PIDxxx |
hbdaq2
|
name server /RUN_switch/RC/ (xxx - any number) |
/RUN_switch/RC/logger |
hbdaq2
|
run logger for the run |
/RUN_switch/erwin |
hbdaq10
|
local error window (external process, but killed at the end of run) |
/RUN_switch/SLB/PIDxxx |
hbdaq2
|
name server /RUN_switch/SLB/ (xxx - any number) |
/RUN_switch/RHP/PIDxxx |
hbdaq2
|
name server /RUN_switch/RHP/ (xxx - any number) |
/RUN_switch/prcmanag |
hbdaq1
|
main process manager for the run |
/RUN_switch/RC/crc |
hbdaq2
|
central run contrl (SMC) process |
Use prcdir GUI to check
and remove the processes on the
corresponding nodes
-
The system is booted but it does not move from EXIT.
-
Go to the expert menu of the run control
pannel and click n "ChkDaughterList".
A tcl/tk window will appear where you can check
the state of the different components in the run. If the ECAL state is
exit, first click "dismiss". (do this
in any case) then, again from the expert menu click "ChkDaughtercfg"
click on the grey button with ECAL on the label. This bring ups another
dialog box. Look for "status_agent". Click on the green
"ACTIVE" button to its right. This should turn the button yellow
and change its label to SKIP. Then click "ChCfg"
then dismiss, then dismiss on the first pop-up. Hopefully you'll see "INITIALIZED"
in
the state field of the run control.
-
During the data taking the input
rate drops to zero, and there are
continous
messages in the error windows:
08:48:10 [FATAL] DIST_0 - send_to_rps: Couldn't
send message of type 0x1c00ff to 1331: (unknown class -1) error 65533
08:48:10 [FATAL] DIST_0 - send_to_rps:
Couldn't send message of type 0x270001 to 1331: (unknown class -1) error
65533
08:48:10 [FATAL] DIST_0 - send_to_rps:
Couldn't send message of type 0x120014 to 1337: (unknown class -1) error
65533
08:48:11 [FATAL] DIST_0 - send_to_rps:
Couldn't send message of type 0x270001 to 1331: (unknown class -1) error
65533
-
One or several SHARC processes died
and the switch is full of undelivered messages. You have to terminate
this run and start a new one.
What to do if the system
stacks in one state machine transition and you cannot
terminate the run:
-
Some transitions are slow, please be patien. Some references:
-
Initialized to configure ( ~ 5 minutes)
-
configure to ready ( ~ 2 minutes)
-
ready to standby
(~ 2 minutes)
-
standby to run.
(~ 2 minutes)
-
run to standby.
(~ 2 minutes)
-
standy to ready.
(~ 2 minutes)
-
Go to the central crate controller
and select "ChkdaughterList" on the "expert"
selection bar. The following window will appear:
Please, write in the logbook
the processes wich are still in transition, and dismiss.
-
Go to the central crate controller and select "ChkdaughterCfg"
on the "expert" selection bar. The following window will appear
Cick on all the green windows
until they turn yellow.
Click on ChgCfg . The state in Central
Run control should change to EXIT.
-
Now you can kill the RUN.
What to do if logger
fails:
Get DMA countfrom
SLBs of a component
-
pause the run.
-
go to a hb-crxx machine.
-
run slbdbg -domain /RUN_xxx/SLB/ -pattern "yyy_*"
eg slbdbg -domain /RUN_switch/SLB/ -pattern "OTR_*" to
get status of all OTR SLBs
Get VDS controlregister
dump
-
pause the run
-
logon to hb-vme34
-
cd to ~hb-vds/fed2
-
run ctrl_dump ee to investigate fed controler
-
run fed_dump to investigate fed's
More
trouble shooting by V.Egorytchev
-
the problem with the
start-up of CRC
GUI can be found here
-
the problem with the
run start-up can be found here
-
the problem with DQ display
can be found here here
EXPERT
INFORMATION.
Where to find the subdetector
specific processes:
SUBDETECTOR |
VME CONTROLLERS |
ECAL Pretrigger |
hb-vme42, hb-vme43 |
FLT |
hb-vme63 |
TRD |
hb-vme58 |
ITR
|
hb-vme59, hb-vme72, hb-vme74
|
SVD |
hb-vme33, hb-vme34,hb-vme57 |
ECAL |
hb-vme45, hb-vme41, hb-vme62 |
OTR |
hb-vme39 |
RICH |
hb-vme36 |
HIPT |
hb-vme49 |
MUON |
hb-vme40 |
EVC |
hb-vme25.1(the one with controller) |
SWITCH |
hb-vme25.1, hb-vme25.2, hb-vme27 |
Where to find the SPC_GUIprocesses:
COMPONENT |
PROCESS |
NODE |
COMMON |
HV_veto |
hb-cr21 |
|
DB_Monitor |
hb-cr21 |
|
4LT_Control |
hbdaq10 |
|
Central_crate_control |
hbdaq10 |
DAQ |
Nodes_Monitor |
hbdaq9 |
TARG |
targ_db_mirror |
hbdaq9 |
SVD |
High_Voltage |
hb-cr26 |
|
Temp_Scan |
hb-cr26 |
|
Chip_Power |
hb-cr26 |
OTR |
Crate_control |
hbdaq9 |
|
Gas_Gain |
hb-cr20 |
|
ASD8_Control |
hbdaq9 |
|
High_Voltage |
hbdaq9 |
ITR |
High_Voltage |
hbdaq9 |
|
High_Voltage_Current |
hbdaq9 |
|
High_Voltage_GEM |
hbdaq9 |
|
Low_Voltage |
hbdaq9 |
|
Gas_Monitor |
hbdaq9 |
|
spc_itr_lhcbvcur |
hb-cr21 |
HighPT |
High_Voltage |
hbdaq9 |
|
Crate_Control |
hbdaq9 |
|
Low_Voltage |
hbdaq9 |
RICH |
High_Voltage |
hbdaq9 |
|
Low_Voltage |
hbdaq9 |
ECAL |
Crate_Control |
hbdaq9 |
MUON |
Crate_Control |
hbdaq9 |
|
High_Voltage |
hbdaq9 |
Where to find the rest of the processes:
NODE |
FUNCTION |
hbdaq2 |
Error logger |
hbdaq2 |
RHP gatherers and savers. |
hbdaq3 |
target poller |
hbdaq2 |
cna_manager |
hbdaq2 |
SMC branch |
hbdaq2 |
snc |
hbdaq2
|
Name server
|
hb-cr17 |
ECAL monitor node |
hbdaq1 |
sub prc managers for the run |
hb-cr10 |
crcui |
hbserv/hera-b/hb2ltlog |
LOGGING |
hb4ltctl |
l2 to l4 controllers |
hb4ltxxx |
4LT |
hbsltxxx |
FLTE/EVA/SLT |
hbslt238 |
TARGET DISTRIBUTOR |
hbdaq0 |
TARGET POLLER, main prcmanager |
hbslt239 |
SLT to 4LT bridge controller. |
LAYOUT FOR VME CRATES DISTRIBUTION IN THE FIRST
FLOOR.
What to do if I don't understand the
instructions on this page
If you feel it could adversely affect the quality of the
data being taken, contact the on-call expert. In any case, send an e-mail
to the responsible for this page with your suggestions for improvements.
Top
Run configuration by experts
CRC GUI has been change
recently to prevent shift crew from doing
mistakes
during run configurations. It means that CRCUI
contains default settings for all HERA-B
standard
runs.
The following CRC
parameters are defined by the default settings:
-
FLT MODE
-
SLT MODE
-
Logging schema:
-
Number of SLT streams
(1-3)
-
Number of 4LT streams
(1-3)
-
Number of loggers (1-3)
To configure a run with the above mentioned parameters,
different from the default settings,
one has to do the following:
-
Configure the run as usual with
standard
run settings
-
Open from the run control
panel the run configuration menu
-
Set the changes you need for
your specific run (see the parameters
mentioned above)
-
Start the run without closing the run
configuration menu
-
Close the run configuration menu if the screen is overcrowded
In case of the run termination, you have to configure
the next run again as described.
Running in case of problems with one of LOGGING
nodes
There are three nodes used for Event
Logging : hb4ltlog,
hb-delfi4, hb-delfi5.
There are employed by the DAQ in the same order. In case
of any problems with hb4ltlog
or hb-delfi4
,
runs have to be configured in such a way that the faulty
node is not used. To configure the run with exluded
hb4ltlog
or hb-delfi4
one has to do the following:
-
After the global run configuration is done, go to
"Expert " -> "Nodes in run" -> "Logging node type" and select the
corresponding
logging type (see figure bellow) (no_delfi4
or no_4ltlog
)
-
If you've restarted CRC GUI
the whole procesdure has to be repeated.
Problems with the Central Network Switch in
201
In case all the windows
on hb-cr10 and hb-cr21
are frosen, you cannot
ping any machine in the Control room, there is most probably
a problem
with the Central Network Switch
in room 201. It 's hanging. And since all the computers in the
West Hall are connectes via this switch, no more
network for the experiment.
To solve the problem, please, follow the instructions
bellow:
-
Go to room 201 and re-plug
the power cable of the switch.
-
IP address of the switch is 131.169.180.10.
It has 12 fiber connector, one of the connectors is not used.
-
Power cable has black colour. It's connectee as the
second to the power extender
, that is on the same level as the switch
-
After you see, that computers in the Control Room
are back to life (one can ping them), you have to execute the script located
~hbshift/recover_vmes/recover_vmes on
hb-cr10
-
After the script is finished, kill
the old runs (slow, switch or repro)
-
Use new GUIs for that
and do "Ongoing run"
-
After the connection to the run is done, use "Expert"->"Clean
the run"
-
To the question "Run Configuration"
enter depending on the run type 'switch'
or 'slow' or 'repro'
-
Re-start the Slow Control.
-
Pay attention to Chip_Power
from SVD
-
If it takes too long
for the state transition INITIALIZED->CONFIGURED,
you have to follow the instructions for 'spsd'
process.
-
You'll find the instructions near hb-cr11
(SVD expert machine)
-
After re-start of spsd,
re-start Chip_Power in the Slow
Control
-
Now you can start the data taking 'switch'or
reprocessing 'repro' run.