STARTING THE SERVERS
The observatory requires a number of continuously running programs
otherwise known as servers or daemons.
There are several levels:
- telco is the highest level server that
reads the database and executes the programs that take the data
- Below it is the
telescope server known as tserver.
tserver
is not actually the server but a program that glues together
the actual device servers into one place
- Below this are the other servers such as
rctel
rstage
rslit
rleach
rmanta
and others
- Finally there is weatherd the weather server daemon
CHECKING THE SERVERS
By running the script show proc
you should
an output like.
rct 1634782 1 0 2022 ? S 0:15 /bin/csh /home/rct/bin/weatherd
rct 264528 1 0 2022 ? S 0:01 bulld
rct 792656 1 0 Jan11 ? Sl 0:32 rleach port=0x1345 debug
rct 792634 1 0 Jan11 ? Sl 2:23 rmanta port=0x1344 camera=guider debug
rct 4128703 1 0 2022 ? Sl 8:33 rslit port=0x1342
rct 4000661 1 0 2022 ? S 0:00 rstage port=0x1343
rct 482774 482773 0 Jan10 ? Sl 0:00 rtel port=0x1341 debug=9 wallace=/home/rct/Documents/Pointing/jan2020.tpt
rct 792664 1 0 Jan11 ? Sl 0:00 tserver port=0x1341 port=0x1342 port=0x1343 port=0x1344 port=0x1345
other processes
rct 1209789 1634782 0 18:38 ? S 0:00 sleep 894
There should be:
rtel
listening on port 0x1341
rslit
listening on port 0x1342
rstage
listening to port 0x1343
telco daemon
tserver
gluing five servers
weatherd
bulld
The tserver
program then cements all these servers
If you see multiple copies of tserver
for long
periods, then the server is hung.
tserver
will fork when carrying out multiple tasks
but these will eventually terminate.
STARTING THE SERVERS
startnet.
You must be logged in as the rct user to run this script.
In theory the startnet script
should be able to restart everything when you invoke it
without arguments.
Sometimes you may have to restart say rstage
then try the following:
killall rstage
startnet
REMOVING THE SEMAPHORES
There is a very occasional problem that the semaphores that
are used to regulate communication with the serial server
fail to release.
This has caused some unexplained hangs in the individual servers.
To clear them execute the following script.
rmsem
You can view the semaphores with the ipcs
command.
REBOOTING THE SERIAL SERVER
If you get a message like unable to connect to to
140.252.58.15 it is possible that the ESR904 terminal
server is hung.
To restart it do the following:
- login as rct
- Use the
show proc
to determine the servers
- kill the
tserver
process
- kill the
rctel, rslit,rstage
servers
- Fire up a brower and connect to http://esr904
- In the window press the
Reset
button
- Answer yes to "are you sure
- Restart the servers as above
Note: the serial server can be rebooted via telnet.
- telnet esr904
- use down arrow to select Configuraation
- Hit Enter key
- Use down arrow key to select Reset
- Hit Enter
- Hit Enter to answer yes
- Wait 30 seconds
Note: If you need to bailout of telnet:
Type: Cntrl^]
To the prompt: telnet> type quit.
RESTARTING LEACH CONTROLLER
The CCD controller has a large number of states that it
can enter that may not be the ones you want.
To reset the controller (I think) execute the following list
of commands.
tx ccd_setup reset tim util pon tset=-100
One very simple test of the controller is to type in the
command.
tx ccd_status
and see if the tchip is reasonable (i.e. around -100C).
RESTARTING THE PMAC
Warning: this procedure is not
for the faint of heart.
It should not be done lightly as it will reset many internal
flags of the motor controller and may take a while to get
things running again.
To reset the pmac like a power off restart type the following quotes and all
echo '$$$' | pputs ether=esr904,4000
You will now have to rehome ALL the motors.
tx home2 x y z fw1 fw2
tx home focus # this takes a LONG time
tx home az
tx home ha dec
Instructions Index