It means the IX locks that requested by the update process is not successful because the same page is owned by the select process . and the select operation is also requested rowlock on the record holding by the

update process .

How can i prevent this scenario happened again ?

1. use rowlock hints in the select statement to prevent the page lock on the entire page ?

2. add additional non cluster index with run_no and desp_status as indexed column ?

3. use snapshot isolation level or serialize isolation level on all the stored procedure involve in deadlock?