Any difference between an array that has grown and one freshly initialised?

Forum History General Discussion
1

This is one for the people who really understand BeyondRAID. Is there a noticeable difference (in capacity, or any other measurable parameter) between a freshly initialised Drobo disk pack consisting of equal capacity drives on the one hand and, on the other, one that has “grown” over time from a smaller capacity to the same capacity through gradual replacement of drives?

My 5N started out with three 1 TB drives that I happened to have spare (1+1+1+0+0). I added a 5 TB drive to each of the two empty slots (1+1+1+5+5). Then I replaced each of the smaller drives with a 5 TB drive, waiting for data relaying to complete in each case (1+1+5+5+5) -> (1+5+5+5+5) at which point I enabled dual drive redundancy (1+5+5+5+5 DDR) -> (5+5+5+5+5 DDR). I’m wondering if my data layout is less than optimal and has more triple mirrors in place of double parity stripes than it would have if I had freshly initialised the 5 x 5 TB array. Triple mirrors are less efficient in terms of capacity than double parity stripes but maybe BeyondRAID optimises them away.

2

hi maybe in the meantime, you might find some helpful info by looking at the patent?

it certainly would be interesting to setup a matrix table, to note down the various space sizes that are available, as well as confirmed pros and cons for each method of upgrade.

3

Hi Paul. I have spent some time reading the patent but it’s so tediously convoluted that it’s quite difficult to follow. There are a couple of training videos on YouTube - about eight people in a classroom and one of Drobo’s engineers in front of a whiteboard - that are quite illuminating, but there isn’t the depth of information to be found. If I had the time I’d do some experimenting myself - after all, I have nothing on my Drobo I don’t have elsewhere. Perhaps it was too much to hope for a reply from someone at Drobo when there are people with real problems that need sorting.