Depending on the performance of the HDD and position of the data on the plattern, the writing speed of the parity varies:

 

 

 

 

I think the write speed could be better by raising the chunk/block size, so the amount of re-positionings of the Read-Write-Head is reduced. What is the actual size? Did the Limeteach developers already tested a bigger size?

 

I’m not very sure on how this works, but by increasing blocksize would that not also mitigate some of the SMR-issues 

 

based on what i know,  with smr a track will be destroyed if the adjacent track is written and thus has to be re-written. 

This kind of is why smr is stil oké for linear writes but not for random writes.  

 

I would feel that file-system and parity calculation tuning. MAY somewhat mitigate the smr-drawback