Make New Entry, Make Followup Entry
User name R. Michaels
Log entry time 11:48:28 on November 7,2010
Entry number 338981
This entry is a followup to: 338941
Followups:
keyword=implemented writing to adaql1I've implemented the scripts that allow writing data to adaql1.
Turning this feature off, so we're back to just adaql2, is easy.
It's just a flag.
The disk is chosen in such a way to spread the data even across
the 4 disks on each computer, i.e.
/adaqlN/dataM where N = 1,2 and M = 1,2,3,4
The deadtime does not seem to be any worse for adaql1 writing.
This deadtime seems to fluctuate in a way I can't predict.
The figs show how things go for run 8446. At first the DAQ wrote
file dat.0 on adaql2, then dat.1 on adaql1, etc. This pattern
depends on history and how full the disks are, but the present
file is always displayed in datamon. I started/stopped a
few short runs to make sure this worked. These runs are all
good production, just short:
Run 8445 - 8451. Then run 8452 will be normal.
I have also implemented the tape copying and deletion on adaql1.
Do NOT attempt to move, copy or (especially) delete files. DO NOT.
FIGURE 1
FIGURE 2
