February 24, 2000
Dear Everyone:
We are still in the throes of extracting data from
our Records
Management software in order to send it to the vendor to be
converted into the new software.
Last Friday, I set up a kind of duty roster for
PC’s. Legacy PC’s can handle
more boxes than GIL PC’s can.
Boris, a legacy PC, can get through about 50,000 boxes in a
“day”. (“Legacy” PC’s are
ones we kept from before we migrated to the GIL machines.
Some functions can’t be done with GIL PC’s, which is another
great reason for changing software products.)
Batman, another
legacy PC, can only get through 20,000 boxes in about the same time
period. GIL PC’s can manage
10,000 boxes. But if you use
a lot of GIL PC’s, you can get a lot of sets of 10,000 boxes; and that
adds up.
Friday evening, I started Boris on the first 50,000
boxes. Then I gave Batman
the next 20,000 boxes (70,000 boxes so far).
Next, I assigned sets of 10,000 boxes to each of 15 GIL PC’s that
were not being used by any real people at the time.
With luck, we would have extracted 220,000 boxes in one night.
That’s roughly one-third of the entire database.
It was my hope that we could get all of the box data extracted
over the 3-day weekend.
My hopes were dashed.
Extracting data is like, you should excuse the expression,
pulling teeth.
Running all those PC’s at the same time basically
caused a traffic jam on the
Information Highway.
They all slowed down to about one-third speed.
I spent a great deal of time in the office on Saturday, waiting
to see if the files were ever going to finish.
Finally, at 4:30, I pulled the plug and aborted all of the
reports. Each PC had gotten
through less than half of their assigned boxes, with the puzzling
exception of Batman.
I regrouped Saturday afternoon and started the next
set of boxes, using the legacy PC’s and only five of the GIL machines.
I also included another legacy PC that lives out in the
warehouse, which I had forgotten was there.
Fewer total files and boxes, but a better chance of completing.
On Sunday, I returned to work and determined that
Batman was still running rings around everyone else, finishing his task
(20,000 boxes) in less than 12 hours.
The warehouse PC, which inexplicably does not have a nickname,
finished his 20,000 in just over 12 hours.
Boris had also finished, but the GIL machines were still running
their files. Rather than
risk slowing them down even further, I decided not to start the legacy
PC’s on another set until I knew how long the GIL PC’s took to finish.
Than answer was:
20 hours to complete 10,000 boxes.
Now I knew how long it would take to run the next set of files
and could plan my visits to the office accordingly.
I started the next set at 3:00 Sunday afternoon.
Knowing that they would all be finished by 1:00 p.m. the next
day, I arranged to have lunch with “Jeannie” on Monday, which was a
holiday.
Now, before you think this was a grinding weekend,
chained to a bunch of PC’s, it was hardly bread-and-water incarceration.
For one thing, there’s a great Chinese restaurant just up the
street. And there’s a
kitchen at work with a microwave oven that even has a “Popcorn” setting
for making microwave popcorn.
Also, the
VCR
and monitor in the conference room (where my five GIL machines were
operating) allowed me to catch up on plenty of deferred viewing while
waiting to see how long GIL machines take to gobble through 10,000
boxes.
I told “Jeannie” to meet me at my place at 2:00,
secure in the knowledge that the GIL machines would complete their work
at 1:00. When I got to the
office at 1:00, the legacy machines were all done, of course.
The GIL machines were still placidly chewing through their 10,000
boxes. Just in case, I
called “Jeannie” to warn her that I might be a teensy bit late.
I got her phone machine and left a message.
Surely, the GIL machines would finish any second now.
At 1:30, they were still going.
I decided that, if they hadn’t finished by 1:45, I
would drive back to my place, pick “Jeannie” up and go back to the
office from there. At 1:45,
I left. At 2:00, I found
“Jeannie” sitting on the patio, reading her book.
When we returned to the office, I discovered that three of the
five GIL machines had finished their reports at 1:46.
Obviously, they were waiting for me to leave the
room. At the rate we’re
going, we should be finished by the end of the week.
Then we wait four weeks to get the data back and start working in
the new system.
Tune in next week for all the squirrelly ways we’re
going to come up with to try and hold everything together for four
mortal weeks.
Love, as always,
Pete
| Previous | Next |