Recently I caught a massive page-split spike on an AlwaysOn Secondary instance that was not mimicked in any way by the active primary. I'm interested in what might cause this on a readable secondary since everything is "read-only".
-
1Did you enabled your secondary replica for read-only workload? Why am I asking? See this.– SqlWorldWideCommented Apr 28, 2017 at 21:30
-
I did, but I would expect the extra row version space to impact the primary and secondary equally .– vesuviousCommented Apr 29, 2017 at 23:21
-
I agree with what you said. Few things to consider in perfmon you do not know what database the splits r happening or kind of split? If you do not have XE setup(more targeted) or read the logs you will get to the bottom of this. 2 resources to consider. First one and 2nd one. There was a blog by Greg LInwood but I cant find it. [This](sqlblog.com/blogs/michael_zilberstein/archive/2009/02/05/… is also useful.– SqlWorldWideCommented May 1, 2017 at 12:28
Add a comment
|