[ https://issues.apache.org/jira/browse/GEODE-8020?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17099418#comment-17099418 ]
ASF subversion and git services commented on GEODE-8020: -------------------------------------------------------- Commit d85dd2341ca5caf46303e5946450a6b2b1e7fc4a in geode's branch refs/heads/support/1.12 from Owen Nichols [ https://gitbox.apache.org/repos/asf?p=geode.git;h=d85dd23 ] Revert "GEODE-8020: buffer corruption in SSL communications (#4994)" This reverts commit 7850daf4f3f7a22dec2c832ad2e563ae6e0a268e. > 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 > Fix For: 1.12.1, 1.13.0 > > > 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 <48509724+miva...@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)