Steve Riesenberg
Steve Riesenberg
@BeauTaapken, thanks for reaching out! `IpAddressMatcher` uses the `java.net.InetAddress` under the covers to resolve the IP address. That class assumes that `null` means the loopback address. I believe this behavior...
@BeauTaapken thanks for the information about your use case. I agree that this behavior is not intuitive. The `IpAddressMatcher` class is used internally in the framework in such a way...
Related gh-11630
See 865bf23ecc8097831a58bc01ca61887bbdfa82e7
Thanks @ShinDongHun1. Do you have any way to quantify the performance improvement of this change? For example, how much faster is it for 1000 requests, etc.? I'm not sure we...
> I'm sorry I'm not good at English, and thank you so much for your reply! > it was a good experience :) No problem, we can try our best...
> umm.. I meant the total time that all requests are processed, not per request Ok. I think it works out about the same either way. It doesn't sound like...
@ShinDongHun1, thanks for your patience. I put together a pretty simple test to measure performance differences with this code change. I tested just the method `cache.saveRequest(request, response)` with 100,000,000 iterations,...
Merged into 5.8.x as 4ff0724c87ee306e23a3758480417024e762915e
Thanks @ugrave. I've looked at the jars available on [repo.spring.io](https://repo.spring.io/milestone/org/springframework/security/spring-security-saml2-service-provider/) and see that this starting showing up in the artifact in version 5.6.0-RC1, though I can't pinpoint what is causing...