To: Observers and CSO Staff

From: Hiro Yoshida (hiro@poliahu.submm.caltech.edu)

Subject: Backend Computer

Now all the login users on hapuna and puuoo will receive a message when the backend computer is booted up and ready for observations. The backend computer sends a UDP broadcast packet (128.171.255.255) at the end of multi-user boot process. The daemon process on hapuna and puuoo listens to the UDP port 9999 for the packet and then notify users by 'wall'. The current message looks something like this:

********************************************
***                                      ***
*** Backend Computer is up and running.  ***
*** Now you can issue AOS command (^_^;) ***
***                                      ***
********************************************

When the backend computer does not work as expected, people tend to rush into the AOS room and press the reset switch. This is generally a BAD idea. In most cases, the backend computer itself (or the operating system kernel) is fine. Try AOS command first, possibly several times, to restart the backend processes. Even if it is really necessary to reboot the backend computer, do not press the reset switch yet. Login to the backend computer as root and type 'reboot -a'. If you can not login to the backend computer, then, well, go ahead and press the reset switch. Note that it will take a few minutes to boot because the file system is being rebuilt. Also note the following.

The most frequently heard problem is that the backend computer looks dead even after it is reset several times. Actually, the operating system is up but in the single-user mode. This is because the system was not shut down properly and the file system needed to be rebuilt. To switch to the multi-user mode, locate the keyboard in the AOS room which is connected to the backend computer and press control-d (logout or exit).

When you see bad frequency or temperature calibration scan, restart the backend processes with AOS/RESTART command.