This website is no longer maintained. Its content may be obsolete. Please visit http://home.cern/ for current CERN information.
Y2K Efforts at CERN
Sverre Jarp IT/DI
First a definition of the problem
To clarify what we mean by the "millennium bug", "year 2000 problem" or "Y2K bug" we have decided to use the definition from the British Standards Institution:
Year 2000 conformity shall mean that neither performance nor functionality
is affected by dates prior to, during and after the year 2000.
Rule 1 No value for current date will cause any interruption in
Rule 2 Date-based functionality must behave consistently for
dates prior to, during and after year 2000.
Rule 3 In all interfaces and data storage, the century in any date
must be specified either explicitly or by unambiguous
algorithms or inferencing rules.
Rule 4 Year 2000 must be recognised as a leap year.
Situation at CERN
Here is a summary of how the problem is being tackled inside the Lab.
M. Metcalf published an initial CNL Article (CNL 228 - "The Millennium Bug") in the autumn of 1997 together with two related articles covering the issues with CERNLIB (DATIME) and ORACLE.
When M. Metcalf left for retirement I accepted to take over the coordination task at CERN. The first sign of activity from my side appeared in Bulletin 98/23, which summed up a presentation given to the Management Board on the 14th of May.
Currently there are activities on several fronts:
- An active group of divisional Y2K coordinators (from divisions AS, DI, PE and SPL), led by
is busily testing CERNís Administrative System and all its procedures on a dedicated test machine. This activity was started in 1997 and Jean initially wrote two very relevant documents: "L'an 2000 - Pourquoi?" and "L'an 2000 - Certification". Currently, phase 1 of this testing has been completed with very encouraging results and phase 2 is under way. The detailed status is covered in a presentation entitled "Y2K certification status in AS" (Powerpoint application).
N.B. All these documents are available from the "ASDB Web server",
in the "2000" directory, at URL:
- Together with the IT/DIS and IT/CO groups, we have documented "safe" levels of operating systems; across Unix, Windows, Mac, and real-time systems. Articles about
had already been published in the Computer Newsletter
- Letters have been mailed to some of our major software/hardware vendor to make them aware of our general concern related to the Y2K problem. The purpose is twofold: get them to supply information and verification tools and secondly make sure that CERN has on record the fact that the Y2K issue has been raised (should there be nasty surprises later). The following companies have been mailed:
IBM, Hewlett-Packard, Compaq (Digital), SGI, Sun, Microsoft, Olivetti, ORACLE, Cadence, Inférence, Matra, Siemens and Elonex. Beyond this list, divisional coordinators have been encouraged to contact directly companies that are of particular relevance to each Division.
- Quarterly Y2K meetings are being held with all the divisional coordinators in order to review issues of concern. Two meetings have been held in 1998, the first one on September 28th and the most recent on November 30th.
- A review of risk management related to Y2K has been done by Lorenz Stampfli/DSU who has also negotiated Y2K coverage with our global insurance company (provided all reasonable measures have been taken by CERN)
- The IT Division is preparing a test setup in the Computer Centre that will be used to investigate problems related to the central services. Please refer to the "Year 2000 co-ordination" presentation (available on the Web) that I recently gave for details.
- The SL Division has established a full list of the areas that need investigation due to the Y2K bug.
has summarised the work in
the "Year 2000
activities in SL Division" Web page.
If you are responsible..
If you are responsible for a service (or a set of services) that may be impacted by the Y2K bug, here is what we recommend you do:
- Make an inventory of all the detailed aspects of the service(s) that may be impacted.
- Try to size up the risks involved ("What might happen if this alarm fails?", "What if this computer fails to operate?") Does the service have to function in January 2000? If not, when does it absolutely have to be operational again?
- Coordinate your efforts with related service providers at CERN ("Who am I sending date-related information to?", "From whom am I getting such information?")
- Make sure you have full support from your Division leader. Keep your DivisionalY2K coordinator fully up-to-date.
- Apply relevant upgrades to (or replace) critical components of commercial nature (Operating systems, Commercial software packages, etc.)
- Apply relevant corrections to all locally developed software/hardware. This is probably the most important area, since nobody else will fix these problems for you.
- Test all the changes both with normal dates and with the clock moved forward to December 1999 so that you roll into year 2000. Do not forget to test the leap year issue as well.
NB: It is prudent to carry out such tests on a specially isolated instance of the data to avoid problems with production.
- Remember that you are dealing with a BUG, not a well-behaved feature. Bugs tend to appear in places where nobody thought they could!
Other information sources
Fortunately CERN does not have to solve all of the Y2K problems in isolation. More and more information is, for instance, becoming available in the Web.
One good resource centre is the year 2000 product compliance list of the company MBS2000 (
http://www.mbs2000.com/) for both Hardware and Software. EDS have established a relatively comprehensive vendor database (vendor2000). There are also references in the initial CNL Article (CNL 228).
The role of the overall Y2K coordinator
Finally one important notice. As the CERN-wide Y2K coordinator I see my role as two-fold:
- Ensure awareness of the overall Y2K problem in all divisions (through the divisional coordinators). Keep CERNís management informed.
- Provide direct/indirect information and tools that help resolve or minimise the Y2K problem. Encourage early testing and implementation of solutions.
NB: Each Division and each service provider remain fully responsible for ensuring that the necessary Y2K solutions are implemented.
Two related bulletins, No_1 and No_2, have been sent to all divisional coordinators and Division Leaders.
Don't hesitate to send comments, suggestions, worries, etc. directly to the author.
For matters related to this article please contact the author.
Last Updated on December 14th, 1998 at 16:30:39
Copyright © CERN 1998 -- European Laboratory for Particle Physics