my application has a dead end where there is a possibility of conflict on three sides:
1) A MySQL database event that runs every minute and runs a couple of update commands in transactions. 2) Background task = runs every second and runs a couple of insert / update commands in a transaction. (Doctrine) 3) Call Main FrontEnd API - based on the incoming request, run the insert / update commands. Due to the high traffic density of requests, this is also a high density. (Doctrine)
So, what I still understand after reading from the Internet that the low level of read uncommitted isolation will help to do the dirty work, which will help minimize deadlocks. But here, in my case, all conflicting UPDATE / INSERT operations .
Will this level of isolation help here at all?
in - EDIT -
Hi, I actually read this, and one of the points is to use a lower level of isolation. However, my main question was about the effect of the read uncommitted isolation level on the update operation , without asking for a solution at a dead end.
source
share