Hi @Riaz,
Here is a known issue that when you process a read query on a secondary replica, the performance might be much slower than the primary replica due to frequent DIRTY_PAGE_TABLE_LOCK waits.
You may take a carefully reference to below links to see if it could solve the issue.
Regards,
Zoe Hui
If the answer is helpful, please click "Accept Answer" and upvote it.