March 31 2017 Friday
Domino NIFNSF update - you probably don’t want to enable it
Update - April 17 2017 - IBM has fixed the issue in 9.0.1 FP8 IF1.
In my last post about NIFNSF, Christian Hensler left this comment:
I couldn't find anything on the internet, so off I went to the Design Partner forum and sure enough there is a post in there from Michael Bourak. Now this is a NDA'd so I'm maybe skirting the rules here, but there is indeed an IBM reproduced issue with performance with NIFNSF. So this AM I did some testing and I was able to reproduce the issue. Based on my testing, on average, the current NIFNSF implementation is twice as slow as non-NIFNSF databases.
So you many not want to implement it just yet.
In my last post about NIFNSF, Christian Hensler left this comment:
I couldn't find anything on the internet, so off I went to the Design Partner forum and sure enough there is a post in there from Michael Bourak. Now this is a NDA'd so I'm maybe skirting the rules here, but there is indeed an IBM reproduced issue with performance with NIFNSF. So this AM I did some testing and I was able to reproduce the issue. Based on my testing, on average, the current NIFNSF implementation is twice as slow as non-NIFNSF databases.
So you many not want to implement it just yet.
Thank you for sharing - was just about to do it this weekend - guess I'll wait :o)