You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@geode.apache.org by "Owen Nichols (Jira)" <ji...@apache.org> on 2020/05/04 23:26:00 UTC
[jira] [Updated] (GEODE-8020) buffer corruption in SSL
communications
[ https://issues.apache.org/jira/browse/GEODE-8020?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Owen Nichols updated GEODE-8020:
--------------------------------
Fix Version/s: (was: 1.12.1)
(was: 1.13.0)
> buffer corruption in SSL communications
> ---------------------------------------
>
> Key: GEODE-8020
> URL: https://issues.apache.org/jira/browse/GEODE-8020
> Project: Geode
> Issue Type: Bug
> Components: membership, messaging
> Reporter: Bruce J Schuchardt
> Assignee: Bruce J Schuchardt
> Priority: Major
>
> When running an application with SSL enabled I ran into a hang with a lost message. The sender had a 15 second ack-wait warning pointing to another server in the cluster. That server had this in its log file at the time the message would have been processed:
> {noformat}
> [info 2020/04/21 11:22:39.437 PDT <P2P message reader for rs-bschuchardt-1053-hydra-client-1(bridgegemfire4_host1_12599:12599)<ec><v1>:41003 unshared ordered uid=354 dom #2 port=55262> tid=0xad] P2P message reader@2580db5f io exception for rs-bschuchardt-1053-hydra-client-1(bridgegemfire4_host1_12599:12599)<ec><v1>:41003@354(GEODE 1.10.0)
> javax.net.ssl.SSLException: bad record MAC
> at sun.security.ssl.Alerts.getSSLException(Alerts.java:214)
> at sun.security.ssl.SSLEngineImpl.fatal(SSLEngineImpl.java:1728)
> at sun.security.ssl.SSLEngineImpl.readRecord(SSLEngineImpl.java:986)
> at sun.security.ssl.SSLEngineImpl.readNetRecord(SSLEngineImpl.java:912)
> at sun.security.ssl.SSLEngineImpl.unwrap(SSLEngineImpl.java:782)
> at javax.net.ssl.SSLEngine.unwrap(SSLEngine.java:626)
> at org.apache.geode.internal.net.NioSslEngine.unwrap(NioSslEngine.java:275)
> at org.apache.geode.internal.tcp.Connection.processInputBuffer(Connection.java:2894)
> at org.apache.geode.internal.tcp.Connection.readMessages(Connection.java:1745)
> at org.apache.geode.internal.tcp.Connection.run(Connection.java:1577)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: javax.crypto.BadPaddingException: bad record MAC
> at sun.security.ssl.InputRecord.decrypt(InputRecord.java:219)
> at sun.security.ssl.EngineInputRecord.decrypt(EngineInputRecord.java:177)
> at sun.security.ssl.SSLEngineImpl.readRecord(SSLEngineImpl.java:979)
> ... 10 more
> {noformat}
> I bisected to see when this problem was introduced and found it was this commit:
> {noformat}
> commit 418d929e3e03185cd6330c828c9b9ed395a76d4b
> Author: Mario Ivanac <48...@users.noreply.github.com>
> Date: Fri Nov 1 20:28:57 2019 +0100
> GEODE-6661: Fixed use of Direct and Non-Direct buffers (#4267)
> - Fixed use of Direct and Non-Direct buffers
> {noformat}
> That commit modified the NioSSLEngine to use a "direct" byte buffer instead of a heap byte buffer. If I revert that one part of the PR the test works okay.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)