Shield.transport.ssl failure

bblank · March 22, 2016, 8:14pm

When attempting to enable ssl on the transport, we are getting the following error in the log file (pointing to an untrusted certificate authority) and the nodes will
not communicate with one another.

[2016-03-22 10:48:17,145][WARN ][shield.transport.netty ] [node01] exception caught on transport layer [[id: 0x098be7f4, /192.168.2.103:52609 => /192.168.2.100:9300]],
closing connection javax.net.ssl.SSLException: Received fatal alert: certificate_unknown at sun.security.ssl.Alerts.getSSLException(Alerts.java:208)

I have added the 3 certificates that make up our trust chain into the keystore for shield, and I also created a truststore.jks and put the
3 certificates in that as well. Any ideas what we are missing?

Location of keystore:
/opt/elasticsearch/share/elasticsearch/config/shield/node01.jks

Location of truststore:
/opt/elasticsearch/share/elasticsearch/config/shield/truststore.jks

Shield Contents of elasticsearch.yml file:

shield.audit.enabled: true
shield.ssl.keystore.path: /opt/elasticsearch/share/elasticsearch/config/shield/node01.jks
shield.ssl.keystore.password:
shield.http.ssl: true

shield.transport.ssl: true
shield.ssl.truststore.path: /opt/elasticsearch/share/elasticsearch/config/shield/truststore.jks
shield.ssl.truststore.password:
shield.ssl.hostname_verification.resolve_name: false

We have tried this with and without the resolve.name and still it doesn't work. We also validated our certicate has the IP address as a SAN in the certificate.

Any ideas?

Thanks,

Bob.

jaymode · March 22, 2016, 10:02pm

Hi Bob,

This definitely seems to be an issue validating the certificate. If you don't mind could you provide the output of keytool -list -keystore file.jks for both the keystore and truststore?

Jay

bblank · March 23, 2016, 3:00pm

Here are the contents of the keystores (obviously some data has been modified to not harm any animals or careers...)

bash-4.1$ keytool -list -keystore node01.jks

Keystore type: JKS
Keystore provider: SUN

Your keystore contains 4 entries

policy, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): 2C:E7:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR:46:5B:DB
issuing, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): 96:A0:84:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR::19:68
root, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): F5:59:64:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR:76:8A:5D
node01.mydomain.com, Mar 17, 2016, PrivateKeyEntry,
Certificate fingerprint (SHA1): 72:2B:A4:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR::A3:E1

bash-4.1$ keytool -list -keystore truststore.jks

Keystore type: JKS
Keystore provider: SUN

Your keystore contains 3 entries

policy, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): 2C:E7:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR:46:5B:DB
issuing, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): 96:A0:84:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR::19:68
root, Mar 17, 2016, trustedCertEntry,
Certificate fingerprint (SHA1): F5:59:64:THESE ARE NOT THE CERTS YOU ARE LOOKING FOR:76:8A:5D

bblank · March 23, 2016, 3:01pm

Here is a more interesting error in the elasticsearch log file... looks like the client may be presenting its simple hostname instead of a fully qualified hostname....

The subject alternate names in our certs do NOT contain just the hostname, only the fully qualified DNS name and IP address:
DNS Name=node01.com
IP Address=192.168.2.100

How can I get the nodes to present their FULLY qualified DNS names when initiation SLL connections on 9300?

ERROR:

[2016-03-23 09:20:07,231][WARN ][shield.transport.netty ] [node03] exception caught on transport layer [[id: 0xa303628c, /192.168.2.103:39301 :> node01/192.168.2.100:9300]], closing connection
javax.net.ssl.SSLHandshakeException: General SSLEngine problem
at sun.security.ssl.Handshaker.checkThrown(Handshaker.java:1431)
at sun.security.ssl.SSLEngineImpl.checkTaskThrown(SSLEngineImpl.java:535)
at sun.security.ssl.SSLEngineImpl.readNetRecord(SSLEngineImpl.java:813)
at sun.security.ssl.SSLEngineImpl.unwrap(SSLEngineImpl.java:781)
at javax.net.ssl.SSLEngine.unwrap(SSLEngine.java:624)
at org.jboss.netty.handler.ssl.SslHandler.unwrap(SslHandler.java:1218)
at org.jboss.netty.handler.ssl.SslHandler.decode(SslHandler.java:852)
at org.jboss.netty.handler.codec.frame.FrameDecoder.callDecode(FrameDecoder.java:425)
at org.jboss.netty.handler.codec.frame.FrameDecoder.messageReceived(FrameDecoder.java:303)
at org.jboss.netty.channel.SimpleChannelUpstreamHandler.handleUpstream(SimpleChannelUpstreamHandler.java:70)
at org.jboss.netty.channel.DefaultChannelPipeline.sendUpstream(DefaultChannelPipeline.java:564)
at org.jboss.netty.channel.DefaultChannelPipeline.sendUpstream(DefaultChannelPipeline.java:559)
at org.jboss.netty.channel.Channels.fireMessageReceived(Channels.java:268)
at org.jboss.netty.channel.Channels.fireMessageReceived(Channels.java:255)
at org.jboss.netty.channel.socket.nio.NioWorker.read(NioWorker.java:88)
at org.jboss.netty.channel.socket.nio.AbstractNioWorker.process(AbstractNioWorker.java:108)
at org.jboss.netty.channel.socket.nio.AbstractNioSelector.run(AbstractNioSelector.java:337)
at org.jboss.netty.channel.socket.nio.AbstractNioWorker.run(AbstractNioWorker.java:89)
at org.jboss.netty.channel.socket.nio.NioWorker.run(NioWorker.java:178)
at org.jboss.netty.util.ThreadRenamingRunnable.run(ThreadRenamingRunnable.java:108)
at org.jboss.netty.util.internal.DeadLockProofWorker$1.run(DeadLockProofWorker.java:42)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: javax.net.ssl.SSLHandshakeException: General SSLEngine problem
at sun.security.ssl.Alerts.getSSLException(Alerts.java:192)
at sun.security.ssl.SSLEngineImpl.fatal(SSLEngineImpl.java:1728)
at sun.security.ssl.Handshaker.fatalSE(Handshaker.java:304)
at sun.security.ssl.Handshaker.fatalSE(Handshaker.java:296)
at sun.security.ssl.ClientHandshaker.serverCertificate(ClientHandshaker.java:1509)
at sun.security.ssl.ClientHandshaker.processMessage(ClientHandshaker.java:216)
at sun.security.ssl.Handshaker.processLoop(Handshaker.java:979)
at sun.security.ssl.Handshaker$1.run(Handshaker.java:919)
at sun.security.ssl.Handshaker$1.run(Handshaker.java:916)
at java.security.AccessController.doPrivileged(Native Method)
at sun.security.ssl.Handshaker$DelegatedTask.run(Handshaker.java:1369)
at org.jboss.netty.handler.ssl.SslHandler.runDelegatedTasks(SslHandler.java:1392)
at org.jboss.netty.handler.ssl.SslHandler.unwrap(SslHandler.java:1255)
... 18 more
Caused by: java.security.cert.CertificateException: No subject alternative DNS name matching node01 found.
at sun.security.util.HostnameChecker.matchDNS(HostnameChecker.java:204)
at sun.security.util.HostnameChecker.match(HostnameChecker.java:95)
at sun.security.ssl.X509TrustManagerImpl.checkIdentity(X509TrustManagerImpl.java:455)
at sun.security.ssl.X509TrustManagerImpl.checkIdentity(X509TrustManagerImpl.java:436)
at sun.security.ssl.X509TrustManagerImpl.checkTrusted(X509TrustManagerImpl.java:252)
at sun.security.ssl.X509TrustManagerImpl.checkServerTrusted(X509TrustManagerImpl.java:136)
at sun.security.ssl.ClientHandshaker.serverCertificate(ClientHandshaker.java:1496)
... 26 more

jaymode · March 23, 2016, 3:40pm

Thanks for the information. The stores look good.

This indicates that node01 is what is being used to connect to the node and it is not available as a SAN entry or the CN. Can you share the network configuration of your nodes? Do you have a SAN for IP addresses and DNS entries in the cert?

bblank · March 23, 2016, 4:58pm

The subject in our certs:
CN=node01.mydomain.com

Subject Alternative Name:
DNS Name=node01.mydomain.com
IP Address=192.168.2.100

Is there a way to get the nodes to present their FULLY qualified DNS names when initiating SSL connections on 9300?

jaymode · March 23, 2016, 5:13pm

Your network configuration for elasticsearch will come into play for that. That's why I asked for the network configuration from the elasticsearch.yml. I believe if you set network.host or network.publish_host elasticsearch will present the name defined there (you can specify the fully qualified name).

bblank · March 23, 2016, 5:23pm

Sorry -

network.host: node01.mydomain.com

bblank · March 23, 2016, 5:37pm

Just tried the network.publish_host: node01.mydomain.com and no difference, same error.

I may have to request new certificates, but I don't want to do that without some confidence it will fix the problem.

jaymode · March 23, 2016, 6:20pm

Lets see if we can avoid that .

Do you use IP addresses in any aspect of configuration? Maybe for discovery (unicast ping hosts)? Is the short name (node01) registered in DNS or a local hosts file.

bblank · March 23, 2016, 7:17pm

discovery.zen.ping.unicast.hosts: ["node01", "node02", "node03"]

DNS has the fully qualified host name and I don't have access to modify the hosts file on the servers. I have spoken with our Certificate team and they are OK with me requesting new certs if that solves the problem.

jaymode · March 23, 2016, 7:33pm

Can you change those to fully qualified names? That should solve the issue.

bblank · March 23, 2016, 8:34pm

The only nodes we have in the ping list are our master or master eligible nodes. Will this solve communication between the non-master nodes as well?

jaymode · March 24, 2016, 10:08am

I think using the FQDNs wherever you specify network configuration or addresses for the cluster to communicate to should solve the issue (if it is the same problem).

bblank · March 25, 2016, 8:31pm

OK - we have it solved...

We added a Subject Alternative Name of just the hostname NOT fully qualified
We added Client Authentication and Server Authentication to the Enhanced Key Usage in the cert
We limited the Cipher algorithms since our JVM doesn't have the unlimited ciphers (yet)

Thanks for your help.

Bob.

jaymode · March 28, 2016, 10:50am

I'm glad to hear you have it solved. One followup for you:

Were you using the default ciphers and saw a log message about this? Are you using OpenJDK from a linux distribution?

If so, it most likely is that your JVM does not enable one of the authentication providers by default. Enabling unlimited ciphers will not resolve this.

bblank · March 28, 2016, 1:05pm

Yes, we were using the default ciphers. Yes, we use OpenJDK. Here is the error we were seeing:
[2016-03-25 10:49:36,210][ERROR][shield.ssl ] [node01] unsupported ciphers [[TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA]] were requested but cannot be used in this JVM. If you are trying to use ciphers
with a key length greater than 128 bits on an Oracle JVM, you will need to install the unlimited strength
JCE policy files. Additionally, please ensure the PKCS11 provider is enabled for your JVM.

shield.ssl.ciphers: [ "TLS_RSA_WITH_AES_128_CBC_SHA256", "TLS_RSA_WITH_AES_128_CBC_SHA" ]

This fixes the unsupported Cipher above - but we do not want this long term. We have to update our JDK to support this cipher. Do you have a good article on OpenJDK to fix this?

Thanks,

Bob.

jaymode · March 28, 2016, 4:42pm

If you are lucky it may be as simple as uncommenting the appropriate line ${java.home}/lib/security/java.security for the SunPKCS11 provider, which may look something like:

#security.provider.10=sun.security.pkcs11.SunPKCS11 ${java.home}/lib/security/nss.cfg

Topic		Replies	Views
Using Shield SSL with the Java client Elasticsearch elastic-stack-security	7	1592	February 9, 2017
Shield 2.3.5 specifcally transport SSL/TLS failing Elasticsearch elastic-stack-security	17	2191	July 6, 2017
Elastic gives error when I enable SSL Elasticsearch elastic-stack-security	1	471	December 14, 2021
Exception caught on transport layer (es6.4.2) Elasticsearch elastic-stack-security	46	16885	February 14, 2019
Elasticsearch cluster chain cert null excpetion Elasticsearch elastic-stack-security	3	416	July 3, 2020

Shield.transport.ssl failure

Related topics