Rebuilding RAID-5 due to one of its disk being replaced & seeing improved performance

Subject: Rebuilding RAID-5 due to one of its disk being replaced & seeing improved performance

Over past weekend I had to rebuild a set of SoftRAID managed RAID-5 file system volumes attached to the office MP6,1 pseudo file server. Other MP6,1s fetch their data from this pseudo file server's RAID-5 across a Thunderbolt Bridge with file sharing option configured to use AFP. The RAID-5 file systems spanned across 8 disk drives.

Prior to the rebuilds the Client MP6,1s were reporting some 500 MBytes/sec reading from the SoftRAID RAID-5 volumes. After the rebuilds (and there were 4 file systems volumes that had to be rebuilt) had completed the Client MP6,1s now showed they were reading data from the pseudo file server at the rates of 700, 800 and sometimes even 900 MBytes/sec.

So what changed to cause this improved data transfer read rate ?????? Was it related to the rebuilds ? Was it a fluke ? I rule out it being a fluke as repeated testing revealed the 700/800/900 MBytes/sec was a solid realistic number.

Maybe the macOS file sharing software was amiss prior to the rebuilds and was using SMB..... but I cannot really believe that either.

No matter, does anyone know why rebuilding a RAID-5 file system volumes would be a reason for providing improved read i/o rates as I described above ?

Rebuilding a RAID-5 after one of its disks is replaced simply updates all the parity data and syncs the various disk partitions to regain RAID-5 redundancy, right ?

Thanks for any responses.....

In hindsight this improvement could be associated with SoftRAID having reported a potential problem with the disk I replaced as shown in attached image.

The improved i/o read rate could I suppose be due to the elimination of the likely substantial error correction rates occurring on the reported disk's potential failure in near future, rather than the affect of the rebuilds.