As you mentioned it's a 32-bit system, what are your memory configurations .. Total memory, AWE, PAE, Lock Pages in Memory, Min&Max Memory ..
32Bits
16GB RAM
/PAE
AWE Enabled
Lock Pages in Memory = SQL Service Account (Domain account)
Min memory 0
Max memory 13310
This configuration is the same as the other 2 nodes in this FCI (Windows 2003 R2)
Check your Index Fragmentation Levels and if you have outdated statistics
I'm using Ola Hallengren's index optimizer. Runs every sunday. CommandLog states all is good. Just rechecked the index fragmentation, looks bad. I'll schedule my old defragmentation script. I think i'm using Ola's job incorrectly. All the commands being logged look correct. Maybe they are not being run.
And run a server side trace to capture any problematic\long-running queries that are triggering these alerts..
Think I found the culprit. That job I rescheduled to stop earlier. The errors came back when it restarted it's schedule. Strange that it only happens late at night.
Step 1 : Record table usage. (We have too many rogue tables floating around. When found renamed)
Step 2 : Record SP usage from sys.dm_exec_query_stats. (No sys.dm_exec_procedure_stats :crying:)
I'll disable the SP step and see what happens.
I'll also run a trace.
Thanks for all the feedback and ideas guys!