INDEX | ORGANISATION | TIME LINE | PEOPLE | STORY | RESEARCH METHOD | COMMENTARY | FINDINGS | HELP | BOTTOM

Unix migration: Migration

' ... It's not actually parallel running really more two week user acceptance testing really. ... what we're not going to do is make sure that exactly the same thing happens on both boxes, but we're going to try and get as close to that as we can, we want to hit the major business issues such as policy renewal and programme renewal and basically just have a lot of people on the database, that's one of the things, stress testing that we've not been able to do in the department. ... what the users will do is they will actually go in pairs, and share PCs and have one as a Unix PC and one as an NLM PC and put as much data as they can into both. Two weeks of testing, they're happy with what sorts of results are coming out and then we'll actually do the migration at the end, the actual data migration. And then it's sort of live. ... I'm pretty happy with that.

We've actually done more of a..one of the tests we've done already is a parallel run, a parallel run actually with the same data, two machines sat side by side and done things like compare what the details are for this policy and how many policies have been retrieved with this criteria and so on and lots of that sort of testing. ... All the testing, the sort of problems we found, usually they existed on both systems. So it's not..I said to the people that were testing, if it happens on the NLM system, I don't care. That's nothing to do with me. But these have been documented and they will be addressed when we have time, but there are a few glitches still in there..there was in fact more problems with the NLM for one reason or another than there were with the Unix one, so it actually got better and some of the..one or two stored procedures have been rewritten because we drew a line on doing further work on the NLM version while we were doing testing for the last couple of weeks, there are now some improvements on the Unix version compared to the NLM. ... '

X068 page 002.02 (tape 06.1.06) Alan Notes/Tape 14/10/94

Social influence:
Technical influence: Tasks: Testing


Recording
Story
Social influence
Technical influence

INDEX | ORGANISATION | TIME LINE | PEOPLE | STORY | RESEARCH METHOD | COMMENTARY | FINDINGS | HELP | TOP

© Clare Tagg 2000