I want to enable client-certificate authentication in my AKS cluster and I have a basic question which I just don't seem to understand. As per the docs, ingress requires the CA certificate to be stored in a secret. My question is: Assuming that I use client-certificates that have been issued by a trusted CA (that's how it works right? CAs issue client-certificates that they sign?), why would a trusted CA give me their CA certificate to be stored in AKS cluster as a secret? Do CAs just hand out their certificates out to public? Isn't that a security issue? (since I can sign client-certificates using that CA certificate)
The CA certificate .crt file doesn't contain the private key. It only contains the public key + certificate information, which is public and can't be used to sign new certificate. You can safely store the ca.crt in a Kubernetes Secret, it only required the private key for the server certificate.
Related
I am trying to secure the client API in API Management using the client certificates.
context.Deployment.Certificates.Any(c => c.Value.Thumbprint == context.Request.Certificate.Thumbprint)
only checks against the certificates stored in "Certificate" directory and not against "CA Certificate"
How do I get all the certificates stored in CA Certificates and then compare against the thumbprint
There is no runtime access to that collection. The purpose of CA certificates is to make sure that SSL certificates and any certificate you call .Validate on can build and validate chain.
Using Openssl, I generated my own CertificateAuthority (CA) and using this CA I generated self signed server certificate and key. This server certificate and key was used to configure Apache HTTPD proxy and Tomcat server which starts successfully and I am able to load application UI and work as expected (All functionality is fine)
Now since it is a self signed server certificate, the browser URL pane shows Certificate error / Not secure everytime. I try to import this certificate into my windows Trusted Root Certification Authorities store using chrome or IE. The certificate import is successful.
Looking at certificate in the Windows Certificate Manager (certmgr.msc) Windows says it "does not have enough information to verify this certificate". When looking at the certificate path, the only certificate that is shown is the certificate itself (with a yellow exclamation mark), and the Certificate status indicates: "The issuer of this certificate could not be found" and checking the certificate name I find that instead of the CA certificate the server certificate is imported. While I can see the certificate on certmgr.msc , but on the browser the imported certificate does not even show under Trusted Root Certification Authorities. The Common Name (CN), SubjectAlternativeName (SAN) etc are all present as expected as I know they are important.
After I manually import the CA certificate (which I already have) separately from the browser, the UI loads with green padlock as expected.
I checked online for suggestions and I can see many threads exists like this0, this1, this2 this3, this4, this5 without much help. None of them have solved this import issue without any manual CA cert import.
What am I missing? With browser certificate import why is server certificate imported and not CA certificate?
What should I do to import the CA certificate directly reading from the server certificate? Should we install the CA certificate manually? If yes, then how to import this CA certificate on a remote machine when I try to access my application from outside the system (remote system browser)? kindly help.
Finally I was able to figure out how to solve this issue. Answering my own question here so that it might help anyone else facing the same issue.
Initially you generate your own CA and and server private key and then using that CA you sign the server certificate. We were using Apache Httpd proxy server and in the configuration along with SSLCertificateFile, SSLCACertificateFile needs to be set to the CA certificate. Upon loading the URL in the browser, "Certificate error" or "Not Secure" will be shown. While clicking on to importing the certificate, under Certificate details you will see CA certificate followed by server certificate. Click to save the CA certificate locally to Trusted Root Certification Authorities (save CA and NOT server certificate) so that you can install this CA (what we created) and once the CA certificate is imported successfully, the connection will be shown as secure with green padlock in the URL bar. This can be done from local or any remote machine accessing the URL.
All certificates that we newly generate and signed by this same CA will be inherently trusted thereafter.
I am trying to establish mutual X.509 certificate authentication between the MQTT client and AWS IoT.
On the X.509 Certificates and AWS IoT page, it is mentioned that one can
reference the CA root certificate in your device code when you connect to AWS IoT. For more information, see the AWS IoT Device SDKs.
However, when I look through the documentation for the Java SDK as well as the API documentation, I see no mention of how to do this. I have downloaded the CA Certificate, but I do not see where I can pass this in when setting up the AWSIoTMqttClient():
this.client = new AWSIotMqttClient(clientEndpoint, clientId, pair.keyStore, pair.keyPassword);
Could someone please point me in the right direction?
I'm setting up certificate authentication for my project using Tomcat. It works ok for command line client such as cURL.
I have many client certificates installed in Chrome browser. Some are using to connect to my site, others are used for different purposes and not relate to my project.
Every time I connect to my site, Chrome presents a list of client certificates for choosing. These are exactly the certificates that I installed and not others. My questions are:
How Chrome knows which client certificates are for a site to present for choosing?
Tomcat stores those client certificates in its trust store. During
SSL hanshake, Tomcat will request for client certificate. Does it
request for some specific certificates that it knows in its trusted
store so that Chrome knows what to show?
The client certificate authentication is ruled in the handshake phase of the SSL/TLS protocol implemented by browsers.
If the server requires a client certificate authentication (it is
optional), send a message to client with the list of the accepted
certificate authorities (CA). Can be void if server accepts any
certificate.
The client select the certificates installed in client keystore which have been issued by any of these CA's, and present the list to user. In case of Chrome, the browser selects the certificates installed by user from the operating system's Key Store.
User choose a certificate, and the client performs a signature with the private key of the certificate over a known data interchanged during handshake.
Only certificates with private key can be selected during step 2. This is the reason by with the browser does not select the certificates of trusted CA's installed in your device. You do not own the private key
I've found that if you connect to a Cloud SQL instance over SSL the CommonName provided in the server's certificate is my-project-123456:myinstance which renders the certificate un-validatable, as the client expects that the CN to be either the hostname or IP.
Every solution to this problem seems to amount to "just disable validation", which is not acceptable to me because:
Why has GCP decided to do everything else correctly, providing a CA cert and client certificates, only to drop the ball on identity validation? By disabling validation you're basically saying "I'm OK with being MITMed at some point".
What about projects where we can't play fast and loose with validation because of PIPA/HIPAA?
What about mySQL clients that don't support turning validation off? eg: All PHP 5.6 mysql libs using mysqlnd prior to the upcoming 5.6.16 release.
Is there any way to make SSL work correctly on Cloud SQL?
One of the reasons for not having the IP address of the instance in the common name of the server certificate is because these IPs can change. What is the IP address of instance A today can be the IP address of instance B tomorrow, because A was deleted, or A decided that it doesn't want the IP address anymore. So, the instance name was decided as being a more unique identification of the instance.
Also, the mysql client libraries by default have hostname verification disabled. http://dev.mysql.com/doc/refman/5.7/en/ssl-options.html
With regards to MITM attacks, it is not possible to MITM attack a Cloud SQL instance because the server certificate and the each of the client certificates are signed by unique self-signed CAs which are never used to sign more than one cert. The server only ever trusts certificates signed by one of these CAs. The reason for using unique CAs per client cert was because MySQL 5.5 did not support certificate revocation lists, and we also did not want to deal CRLs, but wanted to support deletion of client certs.
We will look into ways of supporting SSL for clients which cannot turn off hostname validation. But I cannot promise an ETA on this.
Cloud SQL Team.