search

therealkevinc / SLOB_distribution

A Git repository used only for distributing the official SLOB release.
Other
40 stars 8 forks source link

Redo tests starting from 12cR2 affected by redo optimization #1

Closed MarisElsins closed 5 years ago

MarisElsins commented 6 years ago

Hi Kevin,

Starting with 12cR2 the HEAVY redo mode (and I think the LIGHT too) does not produce the same volumes of redo as it used to before 12cR1 due to optimization I described in https://blog.pythian.com/redo-volume-optimization-in-12c-r2/

I think it'd be useful to modify the SLOB so that the the testing worked around the optimization described. I attempted a quick workaround (https://gitlab.com/MarisElsins/Public/snippets/1728857) and it worked nicely for my tests, but it may not be optimal under more capable systems as I think it might show some contention in obtaining the current SCN. Probably a global PL/SQL variable would be a better option for generating different values to use in the update statements.

P.S. reporting it here as I couldn't find a better way to reach you. P.S.2. I'll remove the snippet from the Gitlab once you've seen it.

Maris

MarisElsins commented 5 years ago

Fixed in SLOB 2.5.0! Thanks, Kevin!

therealkevinc commented 5 years ago

No! Thank you!!

Sent from Yahoo Mail on Android

On Sat, Oct 5, 2019 at 12:01, Maris Elsinsnotifications@github.com wrote:
Fixed in SLOB 2.5.0! Thanks, Kevin!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.