Re: Security Bulletin: IBM WebSphere Application Server Liberty for IBM i is vulnerable to man-in-the-middle spoofing attack (CVE-2022-39161)

First, you should be complete with cover letter:

Here is an example.

To enable integrated web services server WSERVICE to allow
certificates with DNS names of ibm.com, developer.ibm.com:

webPluginConfig.sh -operation *UPDATE
-server WSERVICE -secureHostVerification true
-hostAliases ibm.com,developer.ibm.com

The use of ibm.com,developer.ibm.com is an EXAMPLE.

Second, the “fix” enables you to ensure certificates used for TLS connection that the HTTP server plug-in receives from the application server is what is expected. Domain names or IP addresses are specified in the Subject Alternative Name field of the certificate. It is simply another check.

If you want more detailed information, please read the CVE.

-----------------------------------------------
Nadir Amra
e-mail: amra@xxxxxxxxxx


From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> on behalf of Rob Berendt <robertowenberendt@xxxxxxxxx>
Date: Monday, October 23, 2023 at 12:46 PM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxxxxxxxx>
Subject: [EXTERNAL] Security Bulletin: IBM WebSphere Application Server Liberty for IBM i is vulnerable to man-in-the-middle spoofing attack (CVE-2022-39161)
First, how critical is a CVE that begins with 2022 in which IBM issues a
fix in October of 2023?

Next, does the following statement indicate how much of a sick joke
"service extension" is? <snip>The issue can be fixed by applying a PTF to
IBM i. IBM i releases 7.5, 7.4, and 7.3 will be fixed.</snip>

Now, the real question: When reading the PTF cover letter for 7.5
https://www.ibm.com/support/pages/ptf/SI84088
I see: "To enable integrated web services server WSERVICE to allow
certificates with DNS names of ibm.com, developer.ibm.com". Is this to
restrict what domains can access my webservice? Does one really want to do
that? I'm not an apache configuration expert by any means. Is this just
an additional benefit IBM threw in with this PTF and not related to the CVE?

Is this the part of this PTF which really addresses the CVE?
"If you want to enable hostname certification validation for the
HTTP Web Administration Server, specify *ADMIN for the -server
parameter."

I'm guessing the PTF does nothing to fix the CVE, if you do not turn on the
additional configuration changes to address this CVE.