2
2087
Guest
Ah yes - that was the point of your post wasn't it! I focussed on the idea FORWARD-ONLY improving performance and forgot about the rereadnolock 
I have rerun the tests using -rereadnolock. I can confirm that -rereadnolock slows it down and using FORWARD-ONLY gets it back to the speed (or very nearly) that you would have without -rereadnolock. FORWARD-ONLY appears to make no difference unless -rereadnolock is being used. DO vs REPEAT - I see no difference in speed.
Continue reading...
I have rerun the tests using -rereadnolock. I can confirm that -rereadnolock slows it down and using FORWARD-ONLY gets it back to the speed (or very nearly) that you would have without -rereadnolock. FORWARD-ONLY appears to make no difference unless -rereadnolock is being used. DO vs REPEAT - I see no difference in speed.Continue reading...