For outbound rules for instances that need access to external APIs

[iwata2023]

We use the cloud service AWS.

The following issue has been detected in Prisma Cloud.

> This policy detects security groups where network 0.0.0.0/0 and ::0/0 are used. This policy detects security groups where network 0.0.0.0/0 and ::0/0 are used.
> Select "outbound rules" and click "edit outbound rules"
> Locate rules containing 0.0.0.0/0 or ::/0 and delete them by selecting "x" icon

The ECS assigned to this security group must have access to the Twitter API.

The problem is that the IP of the Twitter API can change.

If the IP of the Twitter API is assigned to a security group outbound, it will be inaccessible when the IP changes.

We have come up with the following solution to this problem.

If ECS tries to access the Twitter API and cannot
The DNS server is queried for the Twitter IP and the outbound security group is updated.

However, this method may not be updated in time and may cause errors in high traffic situations.

Is there another better solution ?

How do you all respond to such alerts ?

[TommyHunt]

@iwata2023 

Given twitter's API endpoint IPs are changing and that twitter is assigned a class A, B or C address range,

Then why don't you use CIDR notation for the AWS Scecrity Group's outbound rule?

See https://en.wikipedia.org/wiki/Classless_Inter-Domain_Routing

As an example, if you enter 199.167.52.5/32, only one IP address is allowed. If you enter 199.167.52.0/24, it allows all IP addresses within the range of 199.167.52.0 to 199.167.52.255.

In that way, you won't have to detect and remediate twitter API endpoint address(es) in your security group(s).  Rather, you trust the range of IP addresses that twitter's API endpoints may occupy.