It looks like a standard question, but I couldn\'t find clear directions anywhere.
I have java code trying to connect to a server with probably self-signed (or expir
If 'they' are using a self-signed certificate it is up to them to take the steps required to make their server usable. Specifically that means providing their certificate to you offline in a trustworthy way. So get them to do that. You then import that into your truststore using the keytool as described in the JSSE Reference Guide. Don't even think about the insecure TrustManager posted here.
EDIT For the benefit of the seventeen (!) downvoters, and numerous commenters below, who clearly have not actually read what I have written here, this is not a jeremiad against self-signed certificates. There is nothing wrong with self-signed certificates when implemented correctly. But, the correct way to implement them is to have the certificate delivered securely via an offline process, rather than via the unauthenticated channel they are going to be used to authenticate. Surely this is obvious? It is certainly obvious to every security-aware organization I have ever worked for, from banks with thousands of branches to my own companies. The client-side code-base 'solution' of trusting all certificates, including self-signed certificates signed by absolutely anybody, or any arbitary body setting itself up as a CA, is ipso facto not secure. It is just playing at security. It is pointless. You are having a private, tamperproof, reply-proof, injection-proof conversation with ... somebody. Anybody. A man in the middle. An impersonator. Anybody. You may as well just use plaintext.