Project Planning with Endeavor
The Imp Team chairs plus selected members of the groups spent Mon and Tues with representatives from Endeavor in a 2-day project planning sessions. Our Endeavor guests were Karen Gegner, Project Mgr Team Leader; Adriana Pilecky-Dekajlo, our Endeavor Project Manager; Jana Fast, our Data Analyst.
We had several good discussions about data migration, location code mapping and indexing. As a result we have a better understanding of the Voyager system. We reviewed a project plan and set many significant milestones. (I will distribute the project plan this afternoon).
Overall the meetings were useful. We came away with an increased sense that we are approaching the implementation in the right way and on a good schedule. I think Endeavor came away with a better understanding of our environment. We were slightly frustrated at times during their presentations because they were generic and did not incorporate our level of sophistication and intention to act as a self-managed site, the suite of products we are buying, the Yale specific contract requirements, or our completed data migration questionnaire. I believe they have a better sense of our plans now.
I was disappointed that milestones were still not known for the Endeavor development issues and for our custom data migration requirements. I emphasized the pressing need for these dates with Karen, Adriana and Jana, followed-up with a call to Tom Owens, Dir. of Implementation, and took advantage of Jane's timely email inquiry about how things were going to gently press her. Verne Coppi, Dir of Development called me the next day (see Development below)!
David reported about the Sun certification test (Aug 20-24).
- Voyager ran successfully on Sun's new Sunfire Ultra Sparc III processor system. (In fact, timings for bulk Marc record loads and index re-builds were very impressive.)
- The tests of SNDR (System Network Data Replicator) and II (Instant Image), 2 SUN products that ITS wants to use to mirror the system on redundant servers to create fail-over capability, also performed well.
- The firmware fix to the T3's (the disks we would use) could not be tested. Sun committed to demonstrating this fix at Yale in early Sept.
Because these tests were so successful and an Ultra Sparc III processor system seems likely, I asked David to prepare preliminary reports to the MMG about our Ultra Sparc III processor-based system configuration options. However, ITS will not be able to commit to a final configuration recommendation until the T3 tests are finished.
David will prepare 3 reports as follows:
- By Mon, 17 Sept, a hardware configuration for production, test, report writer and development servers using Ultra Sparc III processor-based servers. The configuration will include fail-over capability but will not provide for additional domains or redundant backplanes. The report will include prices for Sun equipment, tape sub-system, disk subsystem and networking requirements (i.e., everything we need for the central systems).
- By Mon, 1 Oct, a hardware configuration for production, test, report writer and development servers using Ultra Sparc III processor-based servers that includes fail-over capability and provides for additional domains and redundant backplanes (the "Cadillac" of fail-over capabilities). The configuration will include prices for Sun equipment, tape sub-system, disk subsystem and networking requirements (i.e., everything we need for the central systems).
- By Mon, 15 Oct, a final recommendation from ITS on our production, test, report writer and development servers configuration.
The MMG will need to review the final recommendation before 1 November so that we can place an equipment order at the start of November. I have asked Cesar to schedule MMG meetings during the weeks of Oct 15th, 22nd and 29th for this purpose.
Timeframe for hardware implementation is:
- MMG decision at the end of Oct
- Order hardware Nov
- Hardware and operating system installed Dec-first half of Jan (Yale)
- Voyager and Oracle installed second half of Jan (Endeavor and Yale)
- Full file, all records test data migration and conversion begins on production hardware 1 Feb.
- Linux server received, installed and ready for installation of SFX software thanks to Richard Crane at ITS.
- Final version of license agreement w/ all attachments distributed w/in YUL for final approval and to be forwarded for signatures to Purchasing.
Pre-July 2002 development issues and notes regarding each from my conversation with Verne follow.
Please pay particular attention to those marked with *. I will provide more information about these over the next 2 weeks as I hear back from Verne and in preparation for a MMG discussion.
-Continued progress in the implementation of Unicode support
See LMC/MMG update regarding Unicode task force
-*Refinements to the software to interface to network security protocols such as Kerberos or LDAP.
Verne wanted to hold off on this development until after production. I explained that we need it for patron empowerment and that I did not think our initial requirements will require the level-of-effort to implement that he was anticipating. He will review our requirements document and we'll discuss. I asked Jeff Barnett to prepare a requirements document before the end of Sept.
-*Granular global change (NOTIS glich-like capabilities)
Verne also wanted to delay work on this issue until after July 2002. I explained that a database of our size cannot be maintained without this capability. I asked him to imagine the work that would be required to run global change for the flips from |x to |v across a database of 6-7 million records if we had to match against the entire heading rather than just the subfield. Just w/in the 250,000 GPO records in Orbis, we have over 150,000 headings that need to have the |x to |v flip applied. (I don't know how many of these are duplicate headings). I told Verne that the capability we need by July 2002 is the ability to change bibl record headings (not other Marc record fields) and that we cannot operate w/o this capability.
-Field-level replacements/merge (e.g., ability to load just TOC tags or just non-roman script tags to matching database records).
This was already planned for the next release.
-Improved course reserves processing to eliminate need to have item-in-hand, support for pending and inactive course lists, bulk activation, deactivation and rush recalls.
Sue Crockford-Peters and Audrey drafted a requirements document for reserves modifications and reviewed this document with Mark Sheft from Endeavor. Mark will create an internal requirements document for review at Endeavor. YUL Course reserve staff will review our requirements document. Endeavor reply expected at the end of Sept.
-Pseudo-patron subnote message-like capabilities
Ongoing email exchanges have described our requirements. We are waiting for a proposal from Endeavor.
-Support for the Old Yale classification
Requirements document was sent 6 weeks ago. Endeavor is beginning to work on normalization issues.
-Patron timeout in circulation checkout
Already planned for next release.
-Encryption between Voyager clients and server
Programming already started. I asked Verne for the encryption algorithm.
-Email shortcut from patron and vendor records
Already planned for this release.
-Text under the circulation client icons.
*Verne does not want to implement this change. The circ icons are gifs (images) and if he adds text he will have to create gifs with the text translated to every language they now support. Additionally, the text will be very small since the gifs are small. Pop-up text already displays whenever the mouse moves over the icons. I recommend we drop this requirement in exchange for a guarantee that the Kerberos work will be in place before we go live. The Kerberos development is one of the partnership projects and as such does not have a delivery date specified. We will discuss at the next MMG meeting when we have more information about the other development issues.
-Retention of settings in the OPAC complex search.
Endeavor requested clarification.
Return to the Migration Management Group Site
Go to Orbis2 Implementation Site