MDM Trouble Report for 2009 Mar 18 Telescope: 1.3m Observer(s): varoius (Kemp et al) (Columbia) Instrument: Templeton/direct Problem(s) Encountered: This is a bit of a summary of about a week of 1.3m direct imaging observing. * Mcgraw has crashed 5 times in 7 nights, and 3 times in the last 2 nights. All crashes have involved the console interaction freezing for the user. Observing has been ongoing at each instance, although no single process or operation appears to have been explictly happening during a statistically signification share of the crashes. On most, but not all, occasions the machine cannot be reached over the network. On most, but not all, machine reboots came up okay -- two boot failures failed on setup and configuration of the swap space and required a subsequent reboot attempt to get the machine running which always came up okay. I recall previous filings of this fault, although I don't recall off-hand what maintenance or investigations have happened so far. That said, preliminary checks might include disk integrity and memory integrity checks with the machine dropped to single-user mode. More robust troubleshooting could involve either swapping out the machine with an appropriately-configured spare or reformatting the hard drive and reinstalling the OS, patches, and MDM programs. Other possible issues could seem to potentially include hardware driver issues or network hardware issues. * The telescope aperiodically exhibits oscillations in the north-south axis. These oscillations are not predictable in location or duration and resolution is not accomplished by any specific measure. They also can happen either when tracking an object or when staring straight up at zenith without tracking. The angular amplitude of said oscillations varies but is approximately of order one arcmiute. The frequency of said oscillations varies but is approximately of order one hertz. I know there have been recent issues in the east-west axis, but I don't recall recent issues in the north-south axis. * The dome alignment aperiodically occults the telescope to the extent of extincting objects up to about 0.25 to 0.5 magnitudes at times before corrections are made. These occultations occur in various parts of the sky and do not appear isolated to times at which the telescope is transiting the meridian. The dome encoder has been correct whenever checked and no tweaks have been made during our 7 nights of observing. I do not know whether the old dome-telescope alignment program has been carried over and applied to the new TCS. Therefore, the current model would seem to be incorrect in some places and/or the allowed error in azimuthal angle would seem to be too large at times. * The instructions for stowing the telescope at the end of the evening refer to homing the telescope, which is also a bit of a long process and should perhaps be reserved for times when the TCS has lost knowledge of the telescope position. The short version of this would seem to involve turning off the tracking and then issuing a point ha=0 dec=+31:57:12 command. A button to accomplish this, or at least the second half of this, would seem to be quite useful. * The usual username-password combination for shutting down the machine elegantly did not seem to work on mcgraw and I don't recall sudo reboot or shutdown capability. * The dome send network command operation continues to fail at times, often very frequently, under the new TCS. * The hot water supply pipe for the sink in the 1.3m common area kitchen seems to rattle severely. It is sufficiently bad that one does not want to even use the hot water so as to substantially thwart dish cleaning and sanitary maintenance. Also, if not preexisting, the rattle seems to be at least deconstructing the location in the wall where it protrudes, if not other areas behind the wall that cannot be seen. ------------------------------ Submitted on 2009 Mar 19 [3:05:44]