Feb 8th 2010 - L.Fini @ Mt.Graham

  • Network config verification: It has to be upgraded. Waiting for Jim Davis on wednesday

  • Made a small procedure to gather OS configuration. A copy of OS configuration for both serves before my modifications has been stored.

  • Created user accounts wfs and adsec on both servers.

  • Added the following packages (and some more due to dependencies):
    • python-devel
    • PyQt
    • PyQt-devel
    • pyfits
    • libXpm-devel
    • openmotif
    • openmotif-devel
    • numpy
    • gmp-devel
    • lrzsz
    • readline-devel
    • kdebase
    • kdebase-devel
    • kdelibs
    • kdelibs-devel
    • ncurses-devel
    • xterm

  • IDL 7.1 installed on both servers (license file is attached).
  • Selinux has been disabled (in /etc/selinux/config): it prevented IDL from running

Feb 9th 2010 - L.Fini @ Mt.Graham

  • Added user AOeng to sudoers (and modified sudoers Default)

  • Kit creation step added to main Makefile

  • Kit created and moved to wfsdx (user: AOeng)

  • Created the directory /local/aolog for log files. Owner: AOeng, group: aoacct

  • Created environment setup for user AOeng. It includes definitions needed to build and install the AO Supervisor

  • Done a make: success

  • Done a make install: failure due to conflicts between installation to be done as root and installations to be done as plain user.

Feb 10th 2010 - L.Fini @ Mt.Graham

  • Reorganized the installation procedure (small modifications to two makefiles). Committed everything to SVN and created a new kit

  • Repeated installations from scratch on wfsdx

  • Network layout revised and discussed with Norm. We agreed on a (temporary) layout using two Switch ports for each AO server and a direct connection between the two. We agreed to perform a test to use the default VLAN for data streams between the two servers, so that to avoid the direct connection.

Feb 11th 2010 - L.Fini @ Mt.Graham

  • More work on the installation procedure. Written a prepare procedure to set up the minimal environment to allow installation. Firstly written as shell script, then converted to python to have more control.

  • Jim Davis set up the network but not exactly following the scheme provided. More discussions about this topic.

Feb 12th 2010 - L.Fini @ Mt.Graham

  • Network work. After modification by Davis I did some reconfiguration of ethernet controllers on wfsdx and adsecdx.

  • Communication tests from Control Room to Clean Room. The required VLANS do not communicate, yet. Jim Davis should send a note to network consultant to verify the switch configuration.

Feb 15th 2010 - A.Puglisi & L.Fini @ Mt.Graham

  • Network VLAN was fixed in the morning. Network tests successful with copper wiring.

  • Communication with WFS internal devices tested locally in the lab with copper wiring.

  • Communication with F.O. from wfsdx server in control room also tested.

  • Completed the software configuration of the WFS part by adding the proper configuration and calibration files. A few configuration details modified to adapt to the local environment.

Feb 16th 2010 - A.Puglisi & L.Fini @ Mt.Graham

  • Fixed bug in AdSec Arbitrator code. Submitted to SVN@Arcetri.

  • Writing a procedure to compare directory trees on local and/or remote machines. It'll be used as a tool to maintain alignedment among various installations.

  • Installation procedure revised to separate installation of executables from calibration and configuration data installation.

Feb 17th 2010 - A.Puglisi & L.Fini @ Mt.Graham

  • AdSec switch has been powered by temporary cable and connected to LBTO LAN. It has been reset and its IP reconfigured following directions by LBTO network manager. It must now be properly configured by LBTO staff.

  • The control of the power distribution has been updated with new features allowing remote on/off of the linear power supply.

  • Full config tree reconstruction started

  • A small modification to AOS command interface coded. Yet to be tested.

Feb 18th 2010 - A.Puglisi & L.Fini @ Mt.Graham

  • Some work on AOS going on.

  • Installation tests on adsecdx. A few link problems have been investigated.

  • After the programmed poweroff the Clean Room switch lost configuration. Some time spent to diagnose. Jim Davis fixed this.

  • IDL license files set up to allow fast switch from the server based license to a standalone (arcetri laptop file) license. A chapter about IDL added to Wiki documentation.

Feb 19th 2010 - A.Puglisi & L.Fini @ Mt.Graham

  • The SVN repository @ Tucson has been populated. It will be used to control Mt.Graham development. A full checkout has been done both on wfsdx and on adsecdx

  • conf & calib directories have been moved to adsecdx.

  • More work, in conjunction with Jim Davis from Tucson, to configure the AdSec switch

Feb 20th 2010 - A.Puglisi, L.Fini & M.Xompero @ Mt.Graham

  • /etc/hosts files configured on both adsecdx and wfsdx. See

  • Commissioning related documentation updated.
  • new IP assigned to ADSEC. The unit is fully reachable by adsecdx pc.
  • MX start config adsec sw.

Feb 21th 2010 - M.Xompero @ Mt.Graham

  • MX continue config adsec sw. Fixed configuration problems and reported in TODO list.
  • FL communication test fails.

Feb 26th 2010 - M.Xompero & A.Puglisi @ Mt.Graham

  • Installed Pyro software on adsecdx in preparation for 4D interferometer.
  • Added AutoTrack command to WfsArbitrator to turn on/off ADC/derot tracking. Added corresponding widgets on WfsControl GUI.
  • Added "gimp" package (plus dependencies) to wfsdx to edit GUI images.

Mar 3rd 2010 - M.Xompero & A.Puglisi @ Mt.Graham

  • Switched IDL license file to laptop version because license server is down or unreachable.
  • New WFS low-level status gui, BoardGui, commited to svn
  • Previous days work on rerotator etc. tracking by WfsArbitrator commited to svn
  • Fixed multithreading bugs in Arbitrator GUIs that were causing occasional crashes.
  • Current svn version: 24. Intended to work with both AO2 and AO3 TCS builds, rotator tracking functionality on WFS enabled when working with AO3.

Mar 4th 2010 - M.Xompero, L.Busoni & A.Puglisi @ Mt.Graham

  • Implemented elevation tracking of ASM bias currents into AdSec Arbitrator code.
  • Loop control data (REC matrices, etc) transferred from Arcetri and stored on adsec control computer.
  • Added command aops to show running AdOpt processes
  • Slope transfer test (WFS->ASM) passed.

Mar 5th 2010 - M.Xompero, L.Busoni & A.Puglisi @ Mt.Graham

  • Pyro module for remote interferomenter control has been installed and tested on both the Interferometer PC and the AdSec control computer.

Mar 12th 2010 - M.Xompero, L.Busoni & A.Puglisi @ Mt.Graham

  • Test of Ethernet diagnostic link from W unit to control computer: full diagnostic data up to 800 Hz (minimum required was 330 Hz).

Mar 13th 2010 - M.Xompero, L.Busoni & A.Puglisi @ Mt.Graham

  • Fixes to logging code to avoid the issues of overwritten logs happened in the last days on the WFS side.
  • Added framework of adsec calibration commands to AdSec arbitrator
  • Fixed bug in shutdown code that could affect the adsec when shutting down the WFS and vice-versa.

Mar 14th 2010 - M.Xompero & A.Puglisi @ Mt.Graham

  • Fixed bug in the logging code that was introduced with yesterday's modifications.
  • Added telemetry-style logging for WFS environmental parameters (temperature, humidity, etc), in addition to the standard log file.

Feb 26th 2010 - M.Xompero & A.Puglisi @ Mt.Graham


May 4th 2010 - M.Xompero, L.Fini

  • Modified security level on adsecdx to allow samba clients
  • System installation procedures (prepare.py and select.py) revised

May 5th 2010 - M.Xompero, L.Fini

  • Some enhancements to AOS: added support for flat selection menu in the GUI, added tertiary position, and some seeing variables.
  • Changed the name prefix of DD variables mirrored into RTDB

May 6th 2010 - L.Fini

  • Added support for "lab mode" in arbitrator.
  • Added a variable to allow the display of "lab mode" warning in AOSGUI
  • Secondary swing arm status mirrored into RTDB
  • A complete revision of AdSec arbitrator started.

May 7th 2010 - L.Fini

  • The full day spent in AdSec Arbitrator revision:
    • All the security checks have been gathered in a single routine.
    • More robust management of various cases.
    • Added a check for swing arm deployement
    • Added a Variable to enamble "lab mode": in lab mode all checks are disabled.
    • A small modifcation to the code generator of the FSM (fsmc) to add some housekeeping functions.

May 8th 2010 - L.Fini

  • Continued Arbitrator revision
    • New version of AdSec Arbitrator coded and compiled
    • AO Arbitrator modified to support new command SetZernikes
    • Code moved to adsecdx and wfsdx
  • Small bug fix in Makefile under contrib directory to allow smooth compilation and installation
  • Small bugfix to prepare.py procedure
  • The new version of AO Sup installed as Version V03

May 9th 2010 - L.Fini, M.Xompero

  • Verified setup on adsecdx of the new release
  • Modified the structure of the calib and conf directories
  • Modified the main makefile to correctly install calib directory

May 10th 2010 - L.Fini

  • A complete revision of the mirror safety procedure in the AdSec Arbitrator has been performed.
  • The new version of AdSec Arb (and related GUI) installed on test installation to be ready to exercise it.
  • A bug fixed in the AdSec FSM.
  • The V03 test installation shows a worsening of the open socket problem. MsgD-RTDB must modified (logging more info) to allow debugging this problem.

May 11th 2010 - L.Fini, R.Briguglio

  • Revision of MsgD-RTDB with addition of some logging items. The logging system has been improved to allow better selection of log verbosity.
  • The new version installed on the running installation to gather data related to the open socket problem.
  • We noticed communication problems with the Adam and set up a procedure to diagnose the source of problems: either the Adam itself or a general networking problem (maybe a port of the adsec-switch).

May 12th 2010 - L.Fini, R.Briguglio

  • Some more time spent around the diagnosys of networking problems. Made some procedures to gather data related to network funcitonality
  • The communication with adam1 and adam2 are still unreliable, while other network nodes around (e.g.: the hexapod) seems not to have problems. Due to the nature of the errors (not only dropped packets, but relatively frequent DUP erros) there are likely some problems with the switch ports.
  • Some ehnacements to the AOS and AOSGUI have been coded: a well visible red label is shown iof the AO Supervisor is in "lab mode" (most security checks are disabled). The AOGUI was modified by F.Tribioli.

May 13th 2010 - L.Fini

  • A wrong command on adsec caused the removal of the calib and conf tree of the running system. Some time was spent restoring them from backups and verifying that everything was correct.
  • We noticed that AOS on the telescope cluster dies when the MsgD is shut down. This has never happened on the test bed system and further investigations are needed. It should not be a major problem, for the moment, because the cause of periodic need to restart the MsgD has been solved (see below).
  • The "trunk" version of AOS has been updated with enhacements from version AO4.
  • The new AOSGUI has been tested.
  • A few bugs corrected in the Arbitrator.
  • The recorded data for network tests show that adam1 and adam2 still have communication problems. A test on the unit in Arcetri is planned to verify wether other units show the same behaviour.
  • A long lasting problem with the MsgD (the number of open file descriptor on the sistem was increasing steadily, although erratically) has been solved: it was due to a bug which only happened when any client was trying to connect twice or more times. The bug has been fixed.
  • The procedure to analyze "ping" data has been enhanced.

May 14th 2010 - L.Fini, A. Puglisi

  • Some modification to AOS to help in debugging.
  • The modifications needed for the "binocular" release have been inspected. The code is ready for the switch to binocular mode.
  • In order to allow more sensible off-line tests of the new version of AdSec arbitrator, the code was modified to allow a "simulation" mode. In the meanwhile a few useful features were added to some libraries. The main purpose of this work is to lower the amount of time needed for the test on the running system so that more time ican be allocated for optical tests.
  • First tests with the AdSec arbitrator in simulation mode.
  • A reorganization of the file structure on the AO servers has been discussed and some modifications have been applied.

May 15th 2010 - L.Fini

  • Simulated tests on the AdSec Arbitrator. The shell safety procedures have been thoroughly tested off-line. Some bugs have been fixed and the logging revised to produce useful info at various levels of verbosity.
  • As a support tool a command procedure for thrdtest has been written to simulate various combinations of value for the main envirnoment variables affecting the shell safety: wind speed, elevation and status if the M2 swing arm.

  • The state machine testing has started. A few modifications in the simulation code were necessary to provide correct return status from simulated functions, in order to be able to test many state transitions.

  • While fixing bugs in the code, a review of the log messages has been performed on many software modules.

  • A review of the network tests for the Adam has been done and a small report written. Both Adam systems have network problems, although one of the two behaves quite worse. Further investigation is necessary.

Topic revision: r39 - 16 May 2010, LucaFini
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback