In fact, in mysql, trigger can detect this case. But when I apply this method to Postgresql, it's not ok. Do you know why?
I will follow your suggestion. However imaging we add a "network_id" column (it contain the id of network) to "subnet" tables.. so I need to do what to prevent a record with overlap CIDR and same network_id.?
Currently, I am Openstacker (do you know Openstack) and try to fix bug related to overlap CIDR with concurent request. I have done with Mysql but not Postgresql.
Nguyen Hoai Nam wrote: > Thank you so much for your time. Let me explain my problem: > > 1. There are some conditions which I must follow it. > > Condition 1: the type of "subnet" is varchar(30), > Condition 2: currently, I am about to use trigger and function to detect and prevent CIDR overlapping > (subnet's value) > > 2. What I am doing > > 2.1 Writing a function can detect overlap CIDR (called cidr_overlap). [...] > 2.2 Writing a function for trigger and a trigger as follows: > [...] > - Create trigger: [...] > > > Please don't follow detail syntax. Something like that. Actually, this method is good active in case > of request sequently. But in case of we have two resquest which insert to DB at the same time (it mean > there are concurrent request). Have you ever see this problem, could please give me some advices to > slove it.
A trigger cannot guarantee that the condition will always be satisfied. At the time that the concurrent triggers run the SELECT, everything is fine.
Do use an SQL contraint to enforce a table constraint. If you really need to store "subnet" as varchar (why?), define the constraint like this:
EXCLUDE USING gist (subnet::cidr inet_ops WITH &&)