The observing system is the union of all the computing hardware and software needed to drive the INT and its instruments. The system is distributed across a number of processors that have to be started independently. For the WFC it consists in:
The standard sequence for starting up is:
The controller is mounted on the WFC at the prime focus. This device is made active when the WFC is put on the telescope and is not normally shut down until the next instrument change. Observers should not need to restart the controller and should call out technical support if it is not on-line.
The controller loads its software from PROM when it is powered up, or when the red reset button on its front panel is pushed. Each time the controller restarts, it loses essential information that was downloaded from the system computer. To make the sub-system operational again, reinitialize from the observing-system command-line thus:
cmd MCA RESET
The autoguider computer is in the clip centre below the telescope and the autoguider's CCD controller is mounted on the instrument at the prime focus. The autoguider computer is normally left running the sub-system between nights. Observers can restart the autoguider but should not normally need to. If the need arises, the TO should do the reset.
The autoguider loads its software and comes on-line automatically when the autoguider computer is booted. The best way to do this is walk down to the clip centre, push the red reset button on the processor board in the autoguider computer. The sub-system takes about 60 seconds to reboot and to start the application. After that, you have to reattach the sub-system to the system computer by (1) closing all autoguider GUIs; (2) running startobssys (see below, under Starting the system computer) for the observer's station and the TO's station. You should not need to restart any other parts of the observing system.
Occasionally, it may be appropriate to restart the autoguider CCD controller. Slew the telescope to access park, go up onto the access gantry, and press the black reset button on the front panel of the CCD controller. Then restart the autoguider computer as above.
The telescope computer is in the clip centre below the telescope and is accessed via the TO's X-terminal. It is normal to restart the TCS every night.
Go to the X-terminal and log in to the telescope computer. The node-name of the computer, the account name and the password should be written on the whiteboard behind you. From the menu, select START. The TCS will be loaded and its information display and local interface will display on the X-terminal.
When the TCS is restarted, it loses essential information that is downloaded to it from the system computer. Each time you start the TCS, you must run
startobssyson the system computer (see below).
The DAS computer and its image-acquisition peripheral (the `FOX rack') are mounted in a low cabinet to the right of the bench carrying the system computer. The VT terminal on top of the rack is the console for the DAS computer. The DAS is normally left running between observing nights, except where there has been a change of instrument or detector or some engineering work.
To restart the DAS, log into the DAS console using the account name and password on the whiteboard. Give the command
obssysand choose from the menu an option which matches the instrument and detector of the day and which has the correct version of the observing system; the version to use should be on the whiteboard. Give the command
startobssysto load the sub-system. Some diagnostic information should be displayed, ending with one message of the form
## DASx: I'm ready to server detector `xxx'.for each CCD in use.
The system computer is a SPARCstation in the control room. The observer's `station' is the graphics console of the workstation. To free up the console for daytime use, it is normal for the observer to log out at the end of each night and log in afresh each evening; hence this part of the observing system has to be restarted each night.
Log on to the SPARCstation console using the account name and password on the whiteboard. Give the command
obssysand choose from the menu an option which matches the instrument and detector of the day and which has the correct version of the observing system; the version to use should be on the whiteboard. Give the command
startobssysto load the sub-system. Various displays and local servers will start up, and the client software will locate and initialize the remote sub-systems (TCS, DAS etc.). When you get the `SYS>' prompt back, the system should be ready to go.
If you restart any of the remote sub-systems (i.e. TCS, DAS, autoguider), you will need to re-run startobssys to bring the sub-system back on-line. Similarly, if any of the local displays or server programs crash, re-running startobssys will restore them. You never need to re-run obssys unless you have logged on the system computer.
The system computer is a SPARCstation in the control room. The TO's `station' is the X-terminal embedded in the telescope-control rack. To free up the terminal for daytime use, it is normal for the TO to log out at the end of each night and log in afresh each evening; hence this part of the observing system has to be restarted each night.
At the X terminal, get a telnet window to the system computer using the node name shown on the whiteboard. Log in using the node name, account name and password on the whiteboard. Proceed with obssys and startobssys as described above for the observer's station.
The startobssys command is supposed to be able to work out which bits of the system are already on-line and which need to be reinitialized. However, an error in this logic could reinitialize some areas of the system that have already been set up. This is not a problem, provided that no observing is in progress. For safety, open both the observer's and TO's station in sequence. If you have to re-run startobssys to recover a crashed sub-system, run it on both stations before restarting observing.