Things needed before observing

• laptop (for comfortable viewing of the AO status display) or other workstation.... with VNC installed and working
• observing checklist
• this checklist
• coffee or Diet Pepsi(TM), prosciutto sandwich, wings, Doritos(TM)... your choice.

For Mac or Windows Users

• Rose Finn's hints on running VNC on MacOS   - also available in your brower here. But be sure you follow the basic steps here, especially in cleaning up.   (Thanks, Rose!)
• David Stark's hints on running VNC on Windows   - But be sure you follow the basic steps here, especially in cleaning up.   (Thanks, David!)

Basic protocol for VNC users

• If you are practicing CIMA, heed the instructions on the observing checklist for which machines to use and not use, and to avoid running the data monitoring routines.
• NEVER EVER reset the VNC password!
• Clean up after yourself! See bottom of this file for instructions!

30 minutes before observing

• Call the telescope operator at (+1) 787 878 2612, Ext. 211, and ask for the password for user "dtusr" at "Observer2". Ask also for the status of ALFA (cover off? amplifiers on? all beams working? anything else new?...)

• ssh -X a2010@remote.naic.edu and enter the account password
• ssh dataview and enter the account password once again
• You are now logged into dataview. To facilitate using two screens on the Linux machines, it is best to startup two independent VNC server sessions. Enter the following at the command prompt TWICE:
        vncserver -geometry 1200x900 (or 1250x950 or 1250x1000 depending on desktop)
• You will get a messages detailing what your new X Desktop numbers are - NOTE THE NUMBER AT THE END (e.g., 1 or 5 or 11 or some other small number).
• In another terminal window on YOUR local machine, issue this command to open your desktop:
        vncviewer -shared -via a2010@remote.naic.edu dataview:N (N is the number of the X desktop noted above).
• In yet another terminal window, again on YOUR local machine, issue the same command - just change N to the other server desktop number. In each instance, you will be prompted for two sequential passwords. The first is the normal a2010 account password. The second is the VNCserver password for the server desktops. In the case of using two VNC server sessions, two desktops should pop up. One session can be used to control CIMA, and the other can be used for the waterfall plots and any realtime IDL analysis.
• In another virtual window on your local machine or an a laptop, login to AO using the a2010 account, and start the AO status display and the IF/LO andPointing Monitors:
        ssh -X a2010@remote.naic.edu
        aostatus
        monpnt
        rxview
  These three items are not critical if they crash, therefore, they do not need to be run in the remote desktop. But, they are VERY useful so pay attention to them. They tell you what is **really** happening.
  Note: See Phil's web page for how to access other telescope monitoring windows.

• Keep an eye on the aostatus window, and have the a2010.cat (located in /share/obs4/usr/a2010/a2010.cat) file open for editing in your text editor of choice (pico or vi work well over the network). As soon as the previous observer is out, check the LST, and estimate and add the slew time (usually around 5 minutes). Put this in for the RA in the galfasrc entry and save the file.

Ready to go!

• Call the telescope operator and wait for the ok. Only one session of CIMA can run on observer2 at a given time. Tell the operator the approximate AZ,ZA you want so that he/she can start moving the telescope as soon as the previous program is done. Also ask him/her to set up GALSPECT. Don't start CIMA until the telescope is close to position (once you start CIMA, the telescope motion will halt).
• Use a terminal in one remote desktop to connect to observer2:
  ssh -X dtusr@observer2
  and type in the password provided by the operator.
• Startup CIMA with the command
  cima --X
  and invoke the version we are currently using: the "The 'smart' version of CIMA (Ver 3.1.08 2010-02-01 17:44 "Smart") (Version updated frequently - this version number current as of 2010Feb)
• After typing in a2010 as the project name, your name as the observer name, and your phone number, you will be authorized by the operator.
• Follow the instructions in the A2010 observing checklist - go to Data Taking section, 'Start New CIMA Session' instructions.
• If you have problems (e.g. CIMA does not start, you can't select the receiver, the telescope doesn't move...) contact the telescope operator.

While observing

• Make sure the IF/LO is set up properly; check out the rxview display (which you started up at the beginning). You'd better be observing 100 MHz centered at 1385 MHz!
• Once data-taking has begun, open Mikael's software for data monitoring on the other vncserver window. (If these tasks do not work, try restarting them.)
        alfaobswin & - monitor the power in dB for each beam - leave this window up
        alfamotorwin & - shows ALFA rotation angle - check at beginning of night
        alfabeamwin & - opens alfabeampos widget, which shows ALFA beam positions - check at beginning of night
        alfadatawin & - waterfall, spectral, and total power plots. To see the waterfall plots, click on 'ALFA pattern' and then 'Plot Type' and select 'waterfall' Note that you will not be able to start alfadatawin if the process is already running on dataview (started by one of the previous observers). If this happens, ask the telescope operator to kill the previous process or, if he cannot, to contact Mikael, Arun, or Phil to do so.

• On another terminal, you can start up an IDL session on one of the fast linux machines (e.g. fusion00,aolc1,2,3,4) so that you can run "quicklook" as a backup to the waterfall plots. Note these instructions make use of the aliases in the a2010 account.
        ssh -X a2010@remote.naic.edu
        ssh -X a2010@fusion00 (or aolc1, etc)
        cd /share/a2010/idl/eggidl
        idl
        @wasinit2
        quicklook,'filename'
  where 'filename' is the complete name of the last complete FITS file, e.g. '/share/pserverf.sda3/wappdata/wapp.20080822.a2010.0003.fits'. (The directory path is specified in the log files - look at the previous night's log.) You have the option to display the bandpasses for each beam, both pols of the first record of the file. Not as nice as Mikael's data monitoring, but quick and convenient for remote observers as an extra check or as backup if the data monitoring programs don't work.
  
  Or if you want to look at an arbitrary record, see the file "checkfitsnotes.txt" in the /share/obs4/usr/a2010 directory.

• Keep an eye on the remaining time until the end of the scan on the AO status display. Occasionally it might stop for no reason, meaning that the WAPPs are no longer taking data, and you will receive no error or warning messages from CIMA. If WAPPS fail, see WAPPS Failure Section in A2010 observing checklist.

• Watch the "CIMA Observation Log Display" window, where error and warning messages are displayed. When you see messages concerning the WAPPs (e.g., WAPP(N) DONE adjusting power, WAPP(N) DONE with spectral line observation, etc.), make sure that all four WAPPs are responding. Do not count on CIMA letting you know that one of the WAPPs has stopped working...

• Check periodically that the declination setting (J2000) is correct. For this, look at the monpnt window to be sure that the line starting with Rslt: shows the correct DecJ. Also check the appearance of the array and its coordinates in the alfabeampos widget, which is opened by the aflabeamwin task. If anything looks amiss, discuss with the telescope operator. In particular, be sure that he has given you control of the telescope!

• Restart the WAPPS (1) before TOGS at the beginning of the night; (2) at A2010 startup; and (3) before TOGS at the end of the night.

• Maintain the log - see the A2010 observing checklist.

Finishing Procedures, including VNC cleanup

• Run TOGS if time is allocated for it. See instructions in A2010 observing checklist.
• Exit out of CIMA, followed by exiting out of the observer2 computer. It is important to do this as quickly as possible, so that the next observer can log in.
• Exit out of any IDL sessions and ALFA/CIMA utility programs
• Kill both remote desktops (X in the upper right hand corner).
• Be sure to do this!   In the window connected to dataview on your local machine where you started the VNC server, issue the following command to kill the sessions (you MUST do this): vncserver -kill :N where N is the desktop number. If something goes wrong with this and you get a message indicating that you must kill the process manually, then follow the instructions on the troubleshooting page ("Problems terminating VNC job") which are summarized here:
  • Check what jobs are running on "dataview" via "ps -ecf". If your VNC connection still appears to be running (note its PID=process ID), then you need to kill it using the linux command (while in a window connected to dataview) "kill -9 PID". Run "ps -ecf"; make sure your VNC has disappeared.
• Exit out of dataview and remote.naic.edu
• Carry on with post-observing procedures, summarized in list of tasks to do after each run and A2010 observing checklist . Finish your soda and Doritos as necessary.

Useful links

• The ALFALFA web page at Cornell
• The master list of ALFALFA declination drift assignments
• The Block allocation summary for current observing period
• A2010 Observing cookbook
• A2010 Observing checklist
• A2010 Observational troubleshooting
• Useful aliases for a2010
• Where in the world is...
• Remote Observing with Arecibo web page