just got some parents of 2325

don't know how much of it is people over fetching work for the shutdown.... next few days will sort it out... i know i over fetched for some of my computers

Computers that downloaded predominately parent units are going to be the slowest to return them. I know I see this behavior with my systems when the cache gets larger. The DCF between the types of units is often off by a factor of 3-8.

If they've downloaded a bunch of parents (stocking up for the downtime) then a 2 day cache will hit 12 days as soon as the longer units start finishing. Since parents go out in batches, any system loading up at that time is going to be slow returning them - even past the deadline.

I think we're seeing a lot of children units because the parents went out en mass just before the upgrade and we're getting a tremendous amount of units generated as those come in.

Of course this is all based on watching the activity of my systems, but it's been a fairly accurate predictor of what will be available over the next day or two. From what I can see, it will be a couple more days before we see many parents.

One 2332 and three 2333, still over 95% children and below being downloaded.

I received parent batch 2339 this morning. Then a strange one downloaded - parent batch 2241 (yes, 2241), and not a repair:

Workunit Status  
Project Name:  Help Cure Muscular Dystrophy - Phase 2 
Created:  01/12/2012 18:05:44 
Name:  CMD2_2241-1CF4_B.clustersOccur-2DJ8_A.clustersOccur_0 
Minimum Quorum:  2 
Replication:  2
CMD2_ 2241-1CF4_ B.clustersOccur-2DJ8_ A.clustersOccur_ 0_ 0--  - In Progress  18/01/12 08:35:35 28/01/12 08:35:35  0.00 0.0 / 0.0 
CMD2_ 2241-1CF4_ B.clustersOccur-2DJ8_ A.clustersOccur_ 0_ 1--  - In Progress  18/01/12 08:33:00 28/01/12 08:33:00  0.00 0.0 / 0.0

In the normal sequence, we saw parent batch 2241 on 29 Dec. Perhaps this is a re-injection after fixing an earlier file download error?

Indeed it is. I've now seen parent batches 2244 and 2245. I now wonder if these batches were held back for some reason, partly because that would help to explain the sudden jump forward in batch numbers from 2240 on 29 Dec to 2255 on 30 Dec (snapshots at same time of day, 8.30 am for me in the land of lemon tea drinkers).That looks good. My current estimate is that the parent batches are averaging between 4.25 and 4.5 per day, implying 28 to 30 days left of them (unless the wind changes strength), followed as you say by decreasing availability of offspring and repairs.