[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://bz.apache.org/bugzilla/show_bug.cgi?id=56448

Mark Thomas  changed:

   What|Removed |Added

 Resolution|--- |FIXED
 Status|NEW |RESOLVED

--- Comment #5 from Mark Thomas  ---
It isn't clear exactly when this was fixed but since Aug 2021 when the test was
re-enabled as part of a larger refactoring[1] the tests have been passing
consistently on all platforms.

[1]
https://github.com/apache/tomcat/commit/5c1b14a9e34bd965146c4941742ff6221a5110c8

-- 
You are receiving this mail because:
You are the assignee for the bug.
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://bz.apache.org/bugzilla/show_bug.cgi?id=56448

--- Comment #4 from Rainer Jung  ---
I should have mentioned, hang happens on Solaris Sparc and JVM is version
1.7.0_80.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://bz.apache.org/bugzilla/show_bug.cgi?id=56448

--- Comment #3 from Rainer Jung  ---
Observed a hanging test for NIO2 during 8.0.52 unit test runs on Solaris 10.
Hangs seem to happen sporadically, observed for .32, .43, .47-.52.

Thread dump for 8.0.52 :

[junit] "main" prio=3 tid=0x0002a800 nid=0x2 runnable [0xfdf7d000]
[junit]java.lang.Thread.State: RUNNABLE
[junit] at java.net.SocketInputStream.socketRead0(Native Method)
[junit] at java.net.SocketInputStream.read(SocketInputStream.java:152)
[junit] at java.net.SocketInputStream.read(SocketInputStream.java:122)
[junit] at sun.security.ssl.InputRecord.readFully(InputRecord.java:442)
[junit] at sun.security.ssl.InputRecord.read(InputRecord.java:480)
[junit] at
sun.security.ssl.SSLSocketImpl.readRecord(SSLSocketImpl.java:934)
[junit] - locked <0xe6d5ab68> (a java.lang.Object)
[junit] at
sun.security.ssl.SSLSocketImpl.readDataRecord(SSLSocketImpl.java:891)
[junit] at
sun.security.ssl.AppInputStream.read(AppInputStream.java:102)
[junit] - locked <0xe6fa2b00> (a sun.security.ssl.AppInputStream)
[junit] at sun.nio.cs.StreamDecoder.readBytes(StreamDecoder.java:283)
[junit] at sun.nio.cs.StreamDecoder.implRead(StreamDecoder.java:325)
[junit] at sun.nio.cs.StreamDecoder.read(StreamDecoder.java:177)
[junit] - locked <0xe6fab7b8> (a java.io.InputStreamReader)
[junit] at sun.nio.cs.StreamDecoder.read0(StreamDecoder.java:126)
[junit] - locked <0xe6fab7b8> (a java.io.InputStreamReader)
[junit] at sun.nio.cs.StreamDecoder.read(StreamDecoder.java:112)
[junit] at java.io.InputStreamReader.read(InputStreamReader.java:168)
[junit] at
org.apache.tomcat.util.net.TestSsl.doRequest(TestSsl.java:157)
[junit] at
org.apache.tomcat.util.net.TestSsl.testRenegotiateWorks(TestSsl.java:131)
...

>From the access log one request plus one reneg seems to have worked:

127.0.0.1 - - [07/May/2018:16:17:45 +0200] "GET /tester HTTP/1.1" 200 2
http-nio2-127.0.0.1-auto-3-exec-5 9
127.0.0.1 - - [07/May/2018:16:17:45 +0200] "GET /tester HTTP/1.1" 200 2
http-nio2-127.0.0.1-auto-3-exec-8 7

but then it hanged until killing it (waited for more than 30 minutes before
observing and killing).

Log lines before hang do not give additional insight:

[junit] 07-May-2018 16:17:45.528 INFO [main]
org.apache.catalina.startup.LoggingBaseTest.setUp Starting test case
[testRenegotiateWorks]
[junit] 07-May-2018 16:17:45.545 INFO [main]
org.apache.coyote.AbstractProtocol.init Initializing ProtocolHandler
["http-nio2-127.0.0.1-auto-3"]
[junit] 07-May-2018 16:17:45.626 INFO [main]
org.apache.catalina.core.StandardService.startInternal Starting service Tomcat
[junit] 07-May-2018 16:17:45.627 INFO [main]
org.apache.catalina.core.StandardEngine.startInternal Starting Servlet Engine:
Apache Tomcat/8.0.52
[junit] 07-May-2018 16:17:45.679 INFO [main]
org.apache.coyote.AbstractProtocol.start Starting ProtocolHandler
["http-nio2-127.0.0.1-auto-3-47468"]

Regards,

Rainer

-- 
You are receiving this mail because:
You are the assignee for the bug.
-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://issues.apache.org/bugzilla/show_bug.cgi?id=56448

--- Comment #2 from Mark Thomas ma...@apache.org ---
NIO2 is failing intermittently on OSX. Seems fine on Linux and Windows so far.

-- 
You are receiving this mail because:
You are the assignee for the bug.

-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://issues.apache.org/bugzilla/show_bug.cgi?id=56448

Petr Sumbera petr.sumb...@oracle.com changed:

   What|Removed |Added

 CC||petr.sumb...@oracle.com

-- 
You are receiving this mail because:
You are the assignee for the bug.

-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

[Bug 56448] Implement a robust solution for client initiated SSL renegotiation with NIO

[bugzilla]

https://issues.apache.org/bugzilla/show_bug.cgi?id=56448

--- Comment #1 from Remy Maucherat r...@apache.org ---
I looked at it, but I am still quite lost about why it works or fails.

I don't see how it could work with async IO (like Servlet 3.1). Even blocking
IO could (possibly) be doing concurrent read / write. Is it really mandatory to
have that capability ?

Could you reproduce a failure with NIO2 on some platforms ? It hasn't failed
for me, although the code is the same.

-- 
You are receiving this mail because:
You are the assignee for the bug.

-
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org