<p>
Some things related to disks:
</p>
<p>
IWS (wocam, 135G)
</p>
<p>
We find that the IWS fills up rather quickly. Data from previous nights has to be deleted quite quickly. This is not a problem really, because data is transferred to the offline workstation, and according to Thomas, to some archive machine.
</p>
<p>
Offline workstation (wocoff, 1 TB)
</p>
<p>
I've had to delete many split files in order to be able to continue to work here. The fact that we blow up data volume as we "reduce" data hurts here. I've not worked directly on the USB drive, only copied the raw files there. Actually I'll give this a try today to get a feel for the performance. Thomas says that he has ordered new disks for wocoff, but they are running late, so they may have ended up on a shelf...somewhere...
</p>
<p>
Btw. the performance of wocoff is quite impressive; typically there are several processes happening at the same time on it:
</p>
<p>
1) Data is copied from DHS to wocoff
</p>
<p>
2) I'm using rsync to copy to the USB drive
</p>
<p>
3) I'm reducing data on it, display using skycat (to terminal).
</p>
<p>
4) Someone else e.g. Dietrich is displaying files using skycat (to terminal).
</p>
<p>
5) Thomas sometimes runs his own reduction recipes.
</p>
<p>
All of this works without serious issues.
</p>
<p>
</p>
<p>
USB disk
</p>
<p>
We've accumulated 378GB in 5 days, 75GB per day. I think it's quite possible that we get much more; we haven't done many 10x biases, flat-fields, and work during the day was often technical. <em>I suggest we buy another USB disk</em> just to be on the safe side, even though Evalso seems to work just fine. It can be a cheaper one (less rugged than the LaCie of ~150 euros, and in a smaller package).
</p>
<p>
Ewout
</p>