EECS Instructional Support, University of California at Berkeley
[ ISG home page ] [ who we are ] [ send us email ] [ search ] [ FAQ ]

                        College of Engineering
            EECS Instructional & Electronics Support Groups


								July 15, 2002

              EECS Instructional Computing - Review and Plans
		                Spring 2002

	Recent Improvements
	Current Initiatives
	Budget Priorities 
	Notable Events 

Recent Improvements
  1) 125 Cory: new EE Mixed Signal Systems Lab for CS150 
     Generous grants from National Semiconductor Foundation and Intel provided
     construction, benches, lighting, networking, A/V and 67 computers.  This 
     lab has over 3400 square feet and is used for senior level design courses 
     dealing with embedded systems, digital design, and mixed signal (EECS150, 


  2) 218 Cory: new lab for EE143
     6 new probe stations were designed by ESG staff and build by the ERL
     Machine Shop.  New furniture and clean room facilities are installed.

  3) 119 Cory: new computing lab for CS152

     Instruction purchased 16 high-end DELL Precision 340s ($30,000) for 119
     Cory and purchased additional RAM for 12 existing PCs in 117 Cory 
     ($6,000).  All 28 of these systems now have 1-GB of RAM, which was 
     required for new CS152 software (Xilinx).

  4) 111 Cory: Multimedia Lab

     Instruction purchased 6 new Mac G4s ($15,000), a slide scanner, 2 photo 
     printers ($2,500) and Abobe software ($2,200) for image editing and 
     animation classes (CS39A, CS39J, CS184).  We  purchased additional RAM 
     for 8 existing PCs in 111 Cory ($5,000).  All 14 of these systems now 
     have at least 1-GB of RAM.

  5) UNIX disk quotas raised to 40MB 

     Intel donated a new DELL PowerEdge 2550 with 240GB of disk space and a UPS.
     We deployed that as the "mamba" file server for UNIX home directories.
     This allowed us to increase the disk quotas of UNIX accounts from 20MB per 
     user (or less) to 40MB in January 2002.
     The former "mamba" (a DELL Poweredge 2300) has been re-deployed as an
     administrative server for student IMAP accounts, WEB-based account
     databases and 'grading' utilities.

  6) 4 new UNIX cpu servers

     Intel donated 2 new 1550 Pentium servers, and we purchased a second 1-GHz
     cpu for each.  These are deployed as the UNIX servers "pentagon" and 
     "rhombus", used mainly by CS186 for Postgres and by CS162 and CS169 for
     mySQL and programming in C++ and Java.

     We purchased 2 new Sun Enterprise 420R servers ($40,000).  These 2 
     servers (star.cs and nova.cs) are deployed as both SunRay servers for 271 
     and 273 Soda as well as general UNIX login servers.  They have reduced 
     the overload on our 2 E5000s (quasar.cs and pulsar.cs), which are being 
     used mainly by EE141 for a CAD application, CS162 for the autograder and 
     by CS classes doing programming in C++, Java and SCM.

  7) 349 Davis: new UNIX lab for CS3

     We installed 34 new SunRays ($17,000) in January 2002.  We will install a
     newer SUN server ($12,000) and an LCD projector ($7,000) in June 2002.

  8) 310 Davis: improved UNIX lab for CS61A

     Instruction funded new SunRays and a SUN server, new network and security 
     features, new floor tiles and a paint job in 2001 (about $80,000).  

  9) Pioneering use of wireless devices by CS classes

     Profs Landay and Joseph obtained 50 IBM Thinkpad 570E and 30 HP Omnibook 
     500 wireless laptops for CS169.  Prof Canny obtained Clio wireless 
     handheld computers for CS160.  Instructional staff supports these devices.
     Some details are in /share/b/pub/

Current Initiatives
  1) Provide computing resources for more students without more lab space

     Instructional lab space will not increase, but we must accomodate more
     users because of increasing enrollments and the additional support of 
     CS3 (400+ students) starting in January 2002.

     Analysis of Instructional space usage and needs:

     One way to ease the space crunch is to make it easier for students to
     work from home, dorm rooms and wireless laptops.   Some initiatives:

     - SUN Java CD (
     - EECS software CD  (
     - IS&T "Connecting@Berkeley" CD (
     - downloadable software (
     - provide "ssh" for free, for MS Windows, UNIX and MacOS
     - support for wireless laptops (AirBears) near the labs
     - increase UNIX server capacity for remote logins
     - implement Win2K Terminal Server capacity for remote logins

  2) Support wireless laptops in Instructional labs
     Instruction is working with the campus and departmental network groups
     on plans to deploy AirBears (campus-wide wireless net) in the vicinity
     of our labs (  This access would be for
     all students (CS160 and CS169 have a special EECS-only service).

  3) Decrease Printing Delays and Costs

     Despite the promise of the paperless office, we are printing more that
     ever ($50K/yr in 2001, up from about $31K in 1998).  We have 26 24-ppm 
     printers in our labs and replace at least 3 of them each year.  Still,
     printing delays caused by high usage and technical failures are common.
     The best solution for now is to have redundant printers, so that if one 
     is down another can be found nearby.  

      - 24x7 printing service, requiring only daily spot checks by the staff
      - short print queues (maximum wait of 2 minutes for a print job)
      - purchase 8 new printers per year ($32K)
      - reduce budget allocation back to $35K (about 10% of our budget)

      - impose printer quotas on students to curb inappropriate use 
      - install WEB-based utility for users to check status, clear queues, etc
      - charge printer lab fees to the students to recover toner and paper 
        costs (subject to approval, starting Jan 2003)
      - use budget allocation ($35K) just to buy and repair printers
     We have considered outsourcing the printer maintenance, but the real cost 
     is in consumables (toner and paper). 

  4) Increase Disk Quotas

       We currently provide 40MB of disk space for a UNIX account, which is the
       only permanent account given to an EECS or CS major.  During a semester, 
       students often get additional, temporary 'class' accounts on UNIX (40MB) 
       or Win2K (60MB).  There is also "/home/tmp" surge space that they can 
       use on UNIX for up to one semester.  But it is awkward to share files 
       amongst all these disk areas, and only 40MB is permanent.

       Disk quotas have increased gradually in the last 2 years but students
       still need more.   The last ABET report recommended that we provide
       100MB per student (EECS and CS majors) for their on-line 'portfolios'.
       Within 1 year, we would like to provide 100MB for UNIX accounts.

       - In June 2002, we will add 240GB and a 2-tape 1.8TB library ($25K) 
	 to our UNIX file server so we can double the UNIX quota to 80MB.
       - In FY 2002-2003, we will add 2 more tape drives and more disk space.

Budget Priorities 
  These are notable items because of the cost or current relevance.

  1) 35 PCs for 277 Soda, 199 Cory ($43K)
     Needed to replace 6-yr old PIIs, used by CS162 and other classes for
     database applications and programming in C++, Java and SCM.  We will
     continue to run UNIX in 277 Soda but need PCs so we have the option
     of running Win2K later if needed. 

  2) 1-GB of RAM for each Win2K PC in 125 Cory ($50K)
     Needed to allow EECS150 to run Xilinx and other required software.

  3) 2 2-processor SPARC servers & 50 SunRays for 199 Cory, 275 Soda ($44K)
     Needed to replace 6-yr old Alphas in Cory and 6-yr old PIIs in Soda,
     for EE and CS CAD and programming classes.

  4) 2 2-processor SPARC servers for 310/349 Davis ($24K)
     New servers in Davis will free up a 4-processor SUN E450.  That will be
     re-deployed as a cpu server for large database and 'autograder' uses.

  5) 8 73-GB UltraSCSI disks & 1 30-tape Exabyte Mammoth library ($25K)
     Add to UNIX file server so we can raise user disk quotas to 80-MB.

  6) 4 Pentium servers for UNIX application servers ($22K).

  7) 4 Pentium servers for Win2K application Terminal Servers ($22K)

  8) 2 LCD projectors ($12K)
     One for CS3 (Prof Clancy) in 349 Davis, one for portable use in labs.

  9) 8 wireless ethernet hubs for AirBears ($10K)

  10) Unavoidable infrastructure assessments:
      Instructional contribution to Soda A/C Bypass ($2K)
      Instructional contribution to EECS per-port network fee ($60K/yr?)
      Instructional contribution to IS&T per-port network fee (???)

Notable Events

  Jun 11 - Instructional home dirs and email service were often down in 
	the last week for maintenance

        Home dirs were down Fri June 7, Sat June 8 and several hours on
        Monday June 11 for maintenance.  This was to complete these
        improvements:  installed a larger capacity tape jukebox, replaced
        8 36-GB disks with 8 73-GB disks, upgraded the operating system.
        The result is a doubling of total disk capacity for home directories
        (from 250-GB to 500-GB).  We recently raised user disk quotas to
        40-MB each, and we needed the extra space to accomodate that growth.

        cory.eecs and this WEB server were down Fri/Sat/Sun for maintenance.
        This was during an effort to install additional processors and RAM,
        from 1 250-MHz CPU and 512-MB to 3 250-MHz CPUs and 1.5-GB RAM.  This
        work will continue this week.  The WEB server is running, but logins
        to cory.eecs are blocked.

  May 08 - Recent Mamba crashes

        May 8: (20 minutes) Mamba.cs froze up at about 6:30am and was
        revived at 6:50am.  Same loss of network connection as on Apr 28.

        May 3: (10 minutes) Mamba.cs froze up at about 3:20pm, was
        rebooted and was back in service at 3:30pm.  There were some disk
        I/O error messages but the file checks showed nothing wrong after
        the reboot.   We will watch Mamba closely, including over the

        April 29-30: (33 hours) Please see
        Explanation of the April 30 Mamba crash for a report about this one.

        April 28: (8 hours) Mamba.cs, the Instructional UNIX home directory
        server, became unavailable sometime Sunday morning and was revived
        at 10:30am.  The computer was OK except that it's network connection
        seems to have gone to sleep.

        Side effects of the downtime could include errors from

        - email requiring .forward file or home dir access on Pasteur
        - grading or autograder cronjobs (Mamba couldn't each Pasteur)

        See below for a description of the symptoms you get on our computers
        when the home directory server is down.

  May 02 - Explanation of the April 30 Mamba crash and your LOST+FOUND directory

        The Event

        Mamba.cs, the Instructional UNIX home directory server, crashed at
        9:50am April 29 and was unavailable until 4pm on April 30.  However,
        some users could not login until 9pm.

        The CS186 instructor directory (WEB site, etc) was restored by 11pm.

        Email service on pasteur.eecs was restarted at about 11pm.

        We thank our student and faculty users, who were extremely patient and
        supportive despite the huge loss of service at this critical time in
        the semester.

        LOST+FOUND directories

        A number of files and home directories were not restored to their
        proper names and locations.   When we couldn't determine where the
        files should go, we put them a directory called LOST+FOUND in your
        UNIX home directory.   Please see the LOST+FOUND/README file for an
        explanation of how you can read these files, and ask us for help
        (inst@cory.eecs) if needed.

        What is "Mamba"?

        Mamba is the file server for all EECS Instructional home directories.
        It is a DELL 2550 (Pentium cpu) running Solaris X86 v 2.8, installed
        in our machine room in 287 Soda.  It has this additional hardware:

          Adaptec SCSI-II controller on motherboard (for internal drives)
          Adaptec differential SCSI PCI controller  (for tape library)
          AMI/LSI MegaRAID 3-channel SCSI controller (for RAID disk arrays)
          DELL 220S external RAID disk cabinet with 8 36-GB LVD SCSI disks

        It has been operating well and without major flaws since we installed
        it in August 2001.  We have maintained current patches, driver updates
        and firmware revision levels on the operating system and hardware.

        What caused this disaster?

        This may be of interest just to show what sys admins are doing when
        you can't find them, if for nothing else.

        1. The Mamba server froze up at 9:50am; even the console was dead.
           I (kevinm) was logged in at the console at the time using the X
           Windows server there.  In hindsight, it may have crashed once
           before under the same conditions (local XWindows user at the
           console), but we rarely do that and there is no known bug or
           patch for such a condition.  (Could be a Solaris or a graphics
           card firmware bug?)

        2. I had to turn Mamba and it's disk enclosure off and on again.
           Ideally (as in the past), the computer would have rebooted, run
           file checks and been back in service within about 10 minutes.
           But in this case, it took us 33 hrs to get the file systems back
           on line.

        3. When Mamba started to reboot, 2 of the 8 disks in the external
           RAID cabinet went into "FAILED" mode (red lights, audible alarms).
           We are running RAID 5, which can recover itself from one failed
           disk but not from 2.   But I have seen this condition before, in
           which the RAID controller itself is confused and the disks are
           not really broken, so I reset them to "ONLINE" and let the reboot

        4. Mamba came up on-line, but access to the filesystems was generating
           many error messages to the console.  So we shut it down and started
           to run the UNIX file check ('fsck') command on the filesystems.
           That command displayed many warnings and error messages indicating
           that it was finding files and directories that it could not name or
           locate properly.   Fortunately, it did retain the owners and dates
           of those files.  In the end, these are the things that we later
           moved to your LOST+FOUND directories.

        5. But on Monday morning, we thought this was not good enough, so we
           moved the 8 disks to another computer for further analysis and
           replaced them on Mamba with 8 new, larger (73-GB each) disks that
           we had recently bought.  These larger disks were to be installed
           in Mamba in June so we can raise everyone's disk quotas.

        6. For the rest of Monday, we worked in parallel, trying to repair
           the filesystems on the original disks and also trying to create
           new file systems on the new disks and start full restores to them
           from tape.

        7. The new disks would not 'newfs' and 'fsck' cleanly, so we did not
           trust the filesystems we were creating on them.  So we replaced
           the RAID controller with our spare, installed the latest firmware
           in it and tried again.  We were able to create filesystems (by
           adding the '-o -b32' argument to 'fsck'), and we started restores.
           Then we found that the full restores crashed the system after some
           time.  This happened several times, even on a restore process that
           had run for hours and, we thought, would finish cleanly in the
           middle of the night (Monday).

        8. So the new disks are somehow incompatible with Mamba (Could be a
           Solaris or a RAID controller firmware bug?  Or a conflict between
           the SCSI controllers for the RAID array and for the tape library?)
           We did not want to take the time to debug that while students were
           waiting.  We'll work on that in June.

        9. In the meantime, we were also running a RAID consistency check on
           the original disks, on the former "Mamba", which we have retained
           as backup hardware.  On Tuesday morning, it was clear that we
           should return those disks to the active file server and use them.
           So we ran the original 'fsck' commands (as in paragraph 4 above)
           and let them finish this time.   /home/aa and /home/cc displayed
           the most warnings and ended up with the most LOST+FOUND files,
           but the great majority of each filesystem was intact.   The rest
           of the day on Tuesday was spent running the filesystem checks,
           distributing the LOST+FOUND files to their owners and checking that
           the grading cronjobs and email services would resume correctly.
           I believe that the consistency check that we ran on Monday did
           reduce the number of LOST+FOUND files that we would have had if
           we had just let the 'fsck' commands finish on Monday morning.

        The home directories were mostly available again by 4pm Tuesday.

        (May 2) We are continuing to run full restores, to another system,
        of the home directories as they were at 3am on April 29.  We can
        recover files and directories for users from that.

        What can we do to prevent it in the future?

        Immediately, we are moving the tape functionality to another system
        (and we have disabled XWindows on the Mamba console!).  Once the
        semester is over, we will debug the problems on Mamba.

        We are reviewing several alternatives (and their costs) that would
        give us redundancy and more security against server downtime, such

        1) Fully redundant RAID 1 mirroring attached to 2 fail-over servers.

        2) #1, using single-vendor solution with full service contract so
           there are no driver or firmware inconsistencies and failed hard-
           ware is replaced immediately.

        3) Duplicate the active disks hourly to backup disks on a second
           server, dump the backup disks daily to tapes on a third server.

                                Kevin Mullally
                                Manager, EECS Instructional Support Group
                                378 Cory Hall, UC Berkeley, (510) 643-6141

  May 02 - Logins were blocked to cory.eecs and DECs in 199 Cory for a day

        The user licenses expired on cory.eecs and the DEC Alphas in 199 Cory
        on May 1 and were renewed at 10am on May 2.   During that time, users
        saw the error message

          Too many users logged on already.
          Try again later.

        In the meantime, users could login to other computers such as pulsar.cs
        and the UNIX workstations on the second floor of Soda Hall.

  Jun 02 - Pop on pasteur was discontinued on May 31th

        (Apr 23) We will be ending pop support on mail server as soon as the
        semester comes to a close. We currently offer and will be pushing
        for the use of imap ssl. See /share/b/pub/ for more information.

  Jun 12 - Instructional and other servers in Soda (*.cs) are down during 
 	air conditioning work today

        (Changed from June 11)

        The department has scheduled a shutdown of air conditioning and
        compute server room cooling in Soda Hall on Wednesday, June 12th
        from 8am to 4pm.

        It is paossible that all EECS Instructional computing services in
        Soda, Cory and Davis Halls will be unavailable during that time.

        This is to install a cooling tower bypass system that will keep
        the computer server rooms cool in the event of a general air
        conditioning failure in Soda Hall.  A/C failures have caused
        several computing outages in EECS in the past year.

  Mar 9 - All Instructional computing resources are available again (4:30pm)

        At about 5pm on Thu Mar 7 there was a campus-wide power outage.
        See for information about that.

        These services were down between 5pm on Mar 7 and about 4pm on Mar 9:

        - EECS Instructional UNIX computers and labs
        - Class and student WEB pages on
        - Email via   Incoming email is queued on
          the sending computers so it should be delivered within a day.

        Windows 2000 users in Cory Hall were able to login on Friday, but
        they could not access computers outside of the building, including
        several departmental servers (\\winsww, \\print, etc).

        Power was restored to Cory Hall by 8am on Fri Mar 8.
        Power was restored to Soda Hall at 6:20pm on Fri Mar 8.

        Service was restored to EECS networks and departmental servers by noon
        on Mar 9.  Service was restored to Instructional UNIX servers and labs
        (including Davis Hall) by 4:30pm on Mar 9.

  Feb 11 - Soda and Cory network was down from about 11am until 11:20am

        An unscheduled network change caused Soda and Cory Hall networks to
        be cut off, which also prevented logins on dv310.cs and in 310 and 349

  Feb 11 - dv310.cs was down from noon until 1pm
  Feb 10 - dv310.cs was down from 6pm until 6:50pm

        The dv310.cs server was down for about an hour on Sunday evening and
        again on Monday.  It is the server for the SunRays in 310 Davis and
        349 Davis, so those systems were unusable while dv310.cs was down.

        We have removed 512 MB of new memory that we added last week, which
        seems to be faulty.

        While dv310.cs was down, students logging in over the network (using
        'ssh') could choose another server, including

                pulsar.cs       quasar.cs       star.cs         nova.cs

  Feb 06 - Mamba.cs was down from 1am until 3pm

        Mamba.cs, the Instructional UNIX home directory server, was down
        from 1am to 3pm today due to an intermittent interrupt conflict
        between the new 1-GB interface, the SCSI RAID controller and the
        SCSI tape controller that are in mamba.cs.  We reassigned some
        interrupts and updated a device driver for one of those cards.

        See below for a description of the symptoms you get on our
        computers when the home directory server is down.

  Feb 01 - Mamba.cs was down intermittently on Friday afternoon

        Mamba.cs, the Instructional UNIX home directory server, was down
        at times on Friday afternoon while we attempted to enable a faster
        (1-GB) internet connection to the server.  This would have replaced
        the current (and typical) 100-MB connection.  We'll try again when
        we have determined why it failed, and we'll try not to do it when
        you are logged in!

        See below for a description of the symptoms you get on our
        computers when the home directory server is down.

  Jan 26 - Network was down Saturday night until about 11:30pm

        The network connection to and betweeen and computers was down for several hours on Sat
        Jan 26.  The connection was restored at about 11:30pm.

  Jan 16 - Downtime: mamba.cs, 5pm-6pm

        Mamba.cs will be down from 5pm-6pm today for installation of a new
        disk and a fiber network interface.   Mamba.cs is our UNIX home
        directory server.

        We will also delay incomimg email to pasteur.eecs during that time.

        See below for a description of the symptoms you'll get on our
        computers when the email and home directory server is down.

  Jan 10 - Downtime: MS Windows connections to mamba.cs

        The "samba" server was reinstalled on Mamba on Jan 10.
        From Jan 8-10 you could not connect to your UNIX home
        directory on Mamba from an MS Windows computer.

  Jan 07 - Downtime: this server was down on Jan 8
  Jan 07 - Scheduled server downtime

        Our full dump schedule slipped a bit, and these UNIX servers will
        be down for dumps on Tue Jan 8:

        server:       approx. down time   affects these:
        ------------  ------------------  ----------------------------------
        cory.eecs     9am-2pm             http://inst.eecs server; HSPICE
        po.eecs       9am-2pm             passwd changes, /home/tmp
        torus.cs      1pm-6pm             /home/tmp2; /share/b
        ------------  ------------------  ----------------------------------

        Users were able to login to other servers, including quasar.cs.

        When cory.eecs was down, users were unable to read WEB pages

        From noon-4pm on Jan 7, the Instructional home directory server
        ( was down for system maintenance (we migrated
        the files to a faster machine).

        Symptoms when email and home directories are missing include:

        - "home directory is /" error message when logging in
        - session hangs up if you try to 'ssh' into an Instructional computer
        - unable to read WEB pages from the
        - lots of annoying "NFS timeout" error messages on your screen
        - new email deliveries will be delayed on pasteur.eecs

        We disable email receipt and relaying through pasteur.eecs when the
        home directory server ( is down.  No mail is
        lost.  Computers that send mail queue messages that are not accepted
        by a remote server, and they resend the messages periodically until
        they are received.

  Jan 03 - UNIX logins will fail because of server maintenance, Jan3-7

  Instructional servers will be down for dumping and maintenance.
  UNIX logins will fail when some of these servers are down:

  server:       approx. down time             affects these:
  ------------  ----------------------------  -----------------------------------
  quasar, star  Wed 1/02             9am-6pm  271 Soda; MMI (ee141)
  pulsar, nova  Wed 1/02             9pm-6pm  273 Soda
  torus.cs      Wed 1/02             9am-6pm  /home/tmp2; /share/b
  SCOTLAND      Wed 1/02             9am-6pm  IIS server; NT class accounts

  pasteur.eecs  Thu 1/03             9am-6pm  Instructional email server
  mingus.eecs   Thu 1/03             9am-6pm  Cadence, CAD WEB site
  mamba.cs      Thu 1/03 & Fri 1/04  9am-6pm  UNIX home directories*
  saidar.eecs   Thu 1/03 & Fri 1/04  9am-6pm  /share/b, UNIX home directories*
  FILESERVICE   Thu 1/03 & Fri 1/04  9am-6pm  NT home directories

  po.eecs       Mon 1/07             9am-6pm  passwd changes, /home/tmp
  cory.eecs     Mon 1/07             9am-6pm  www-inst.eecs HTTP server; HSPICE
  dv310.cs      Mon 1/07             9pm-6pm  310 Davis
  C119Z         Mon 1/07             9am-6pm  NT WORKVIEW (cs152); ee290t
  ------------  ----------------------------  -----------------------------------
                * Type 'quota' to determine which
                  file server has your home directory.

  Jan 2002 - EECS Instructional CD and SUN Java CD are now available

        Conditions of use:
        * student is
          - an EE or CS major
          - OR is enrolled in an EE or CS class
          - OR is authorized for cardkey access to an EECS Instructional lab
        * student uses software for EECS coursework, not for commercial gain
        * only one of each CD per student please.  NOT FOR RESALE.

        Instructional CD:
        The "EECS Instructional Software CD" version 1.0 is now available from
        387 Soda and the CS SelfPaced Center.  This CD contains a collection
        of public domain software used in EE and CS couses.  It is provided
        free by EECS Instruction for students to use on their home computers.
        (It replaces the floppy diskette that was available at 387 Soda.)

        Students in EECS classes can also download the Instructional CD software
        from a new WEB site, prepared by EECS Instruction (thanks, brg!).
        There is an image of the entire site that you can download and burn
        onto a CD, or you can download individual software packages.  See for details.

        Software included (for MS Windows, Linux and/or MacOS):
        ----------------------------    ---------------------------------------
        Adobe Acrobat Reader            Apple  Macintosh Programmers' Workshop
        SCM, DrScheme, MacGambit        Ghostscript
        Emacs                           Java Development Kit (download only)
        MacGzip                         Stuffit Expander for Macintosh
        MIPS development tools          Cygwin for Windows
        Netscape                        Microsoft Windows Installer
        Perl, MacPerl, ActivePerl       eXceed (download only), MacX
        STk                             Info-ZIP for Windows
        SPIM                            Tera Term SSH
        ----------------------------    ---------------------------------------

        SUN Java CD:   (4 CDs)
        The "SUN Java Jumpstart CD" set is also available from 387 Soda and the
        CS SelfPaced Center.  This CD contains a collection of Java programming
        utilities.  SUN Microsystems has donated these CDs to EECS for students
        to use on their home computers.

        Software included (for MS Windows, Solaris SPARC & X86, Linux):
        JavaTM  Development Kit Family:
                JavaTM   2   SDK,  Standard Edition v 1.3
                JavaTM   2   SDK,  Standard Edition v 1.2.2 [Downloads]
                JavaTM   1.1.8  -  Documentation
        Other JavaTM Products and APIs:
                JavaTM   Access Bridge 1.0
                SDK, JRE, JavaBeans, Java Media, APIs, Jini, Jiro, code samples

For additional information, please contact us:
  Kevin Mullally, ISG Manager		|  Ferenc Kovac, ESG Manager
  EECS Instructional Support Group	|  EECS Electronics Support Group
  378 Cory Hall, (510) 643-6141		|  380 Cory Hall, (510) 642-6952		|
					|	|	
  UNIX, Win2K computers and software	| Win2K computers, software and
  in drop-in labs; email and UNIX 	| equipment in electronics labs,
  accounts; UNIX login servers;		| AV services.
  class and student home pages.		|

source: ~iesg/public_html/reports/managers/Spring_2002