kohsuke
08d211bc9e
added error diagnostics to find out where the connection is closed.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@5434 71c3de6d-444a-0410-be80-ed276b4c234a
2007-10-25 00:48:52 +00:00
kohsuke
4afa2a96fc
when a thread that's waiting for a remote operation is interrupted, also interrupt the remote thread that's executing it, so that the actual operation gets cancelled.
...
(#904 )
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@5393 71c3de6d-444a-0410-be80-ed276b4c234a
2007-10-21 23:39:03 +00:00
kohsuke
605ae5d355
missed the crucial part.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4270 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-15 06:02:15 +00:00
kohsuke
5d96a35cab
there are cases where automatic unexporting is undesirable.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4269 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-15 05:59:15 +00:00
kohsuke
cc92e3983d
bug fix.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4258 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-13 22:12:16 +00:00
kohsuke
ab8522ef4a
with automatic unexport operations, the auto close can cause NPE.
...
In any case I think this was a bad behavior for remote streams to cloes the underlying Output/InputStream when it is GC-ed.
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4240 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-12 23:40:53 +00:00
kohsuke
b34e2785f6
improved error handling.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4239 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-12 23:33:24 +00:00
kohsuke
13d58a5d4c
bug fix.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4207 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-11 19:19:41 +00:00
kohsuke
d7c8b59223
bug fix.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4206 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-11 19:19:29 +00:00
kohsuke
573dd39220
implemented reference counting and automatic unexporting at the end of synchronous calls.
...
This is to fix http://www.nabble.com/Possible-memory-leak-in-hudson.remoting.ExportTable-tf4218077.html
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4202 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-11 19:09:12 +00:00
kohsuke
f31f0dcc87
implemented reference counting.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4201 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-11 18:48:09 +00:00
kohsuke
3d92fc3acc
added dump command.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4153 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-09 05:20:01 +00:00
kohsuke
fa209b3ff5
added probe points to dig into the leak problem.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@4152 71c3de6d-444a-0410-be80-ed276b4c234a
2007-08-09 05:11:00 +00:00
kohsuke
99114bf563
improved error check.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@3723 71c3de6d-444a-0410-be80-ed276b4c234a
2007-07-16 22:48:14 +00:00
kohsuke
7321aba356
forgot to commit.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@3675 71c3de6d-444a-0410-be80-ed276b4c234a
2007-07-13 00:39:04 +00:00
kohsuke
38f4b8ba4e
preventive measure to forestall memory leak possibility.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@3673 71c3de6d-444a-0410-be80-ed276b4c234a
2007-07-12 14:03:34 +00:00
kohsuke
babc31f6f7
reimplemented the shut down sequence to avoid EOFException race condition. ( #595 )
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@3668 71c3de6d-444a-0410-be80-ed276b4c234a
2007-07-12 13:10:53 +00:00
kohsuke
75ab95fed3
made protected to allow subclass to perform additional actions (such as killing a process.)
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@3187 71c3de6d-444a-0410-be80-ed276b4c234a
2007-04-21 14:28:19 +00:00
kohsuke
3251761d28
fixed a bug when a stream is closed immediately after it's opened.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2894 71c3de6d-444a-0410-be80-ed276b4c234a
2007-04-06 17:54:02 +00:00
kohsuke
b661e27787
re-implemented correctly.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2801 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-30 06:35:01 +00:00
kohsuke
5ff31c3939
support rt.jar
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2779 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-29 17:00:09 +00:00
kohsuke
21c77af00c
modified so that exception happened on the remote system can be propagated (to some extent) even if it's not serializable.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2756 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-29 00:37:55 +00:00
kohsuke
d4850ef00c
touching.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2656 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-25 22:42:51 +00:00
kohsuke
8eeeb69e1a
improve diagnosability.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2571 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-22 23:25:34 +00:00
kohsuke
8820c0208c
fixed a dead lock. Request.call() shouldn't block in wait() while keeping the Channel lock.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2350 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-06 19:44:15 +00:00
kohsuke
09388df912
fixed a dead lock.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2345 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-05 17:53:27 +00:00
kohsuke
a92734e8ab
support flush.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2323 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-04 04:37:16 +00:00
kohsuke
7581341b30
buffering causes deadlocks
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2322 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-04 02:10:38 +00:00
kohsuke
400bbdaf5b
moved the code to locate the source jar file.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2291 71c3de6d-444a-0410-be80-ed276b4c234a
2007-03-01 05:37:02 +00:00
kohsuke
696c4f1ac0
improved the error reporting.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2242 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-22 01:43:24 +00:00
kohsuke
64dd2b64b5
refactored for better reusability.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2241 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-22 01:42:11 +00:00
kohsuke
8aac70529d
there was apparently some kind of race condition, because occasionally test fails trying to call close once too many (that is, one channel (X) sends CloseCommand to the other (Y), which sends back CloseCommand to X, and somehow X sends one more CloseCommand to Y, causing an exception below.)
...
I thought the use of the 'close' field and synchronization is enough to guard against this condition, but apparently it's not. I added one more 'closing' field, which seems to fix the problem.
It just doesn't feel right when you fix the issue without getting to the bottom, but the concurrency bug is hard to reproduce and I'm exhausted.
----
forking a new process
Feb 15, 2007 6:09:30 AM hudson.remoting.Channel$CloseCommand execute
SEVERE: close command failed
java.io.IOException: Pipe closed
at java.io.PipedInputStream.checkStateForReceive(PipedInputStream.java:190)
at java.io.PipedInputStream.receive(PipedInputStream.java:156)
at java.io.PipedOutputStream.write(PipedOutputStream.java:129)
at java.io.ObjectOutputStream$BlockDataOutputStream.drain(ObjectOutputStream.java:1682)
at java.io.ObjectOutputStream$BlockDataOutputStream.setBlockDataMode(ObjectOutputStream.java:1591)
at java.io.ObjectOutputStream.writeNonProxyDesc(ObjectOutputStream.java:1173)
at java.io.ObjectOutputStream.writeClassDesc(ObjectOutputStream.java:1127)
at java.io.ObjectOutputStream.writeOrdinaryObject(ObjectOutputStream.java:1284)
at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1079)
at java.io.ObjectOutputStream.writeFatalException(ObjectOutputStream.java:1398)
at java.io.ObjectOutputStream.writeObject(ObjectOutputStream.java:305)
at hudson.remoting.Channel.send(Channel.java:183)
at hudson.remoting.Channel.close(Channel.java:347)
at hudson.remoting.Channel$CloseCommand.execute(Channel.java:329)
at hudson.remoting.Channel$ReaderThread.run(Channel.java:372)
Feb 15, 2007 6:09:30 AM hudson.remoting.Channel$ReaderThread run
SEVERE: I/O error in channel
java.io.EOFException
at java.io.ObjectInputStream$BlockDataInputStream.peekByte(ObjectInputStream.java:2498)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1273)
at java.io.ObjectInputStream.readObject(ObjectInputStream.java:348)
at hudson.remoting.Channel$ReaderThread.run(Channel.java:366)
channel started
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2217 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 15:06:04 +00:00
kohsuke
253d7732ee
added code to track the origin of the call across remote systems.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2216 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 15:03:13 +00:00
kohsuke
cec26f9009
trying to mimic the UserRequest class to see if that works.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2215 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 14:11:36 +00:00
kohsuke
274ea9c68e
fixed a bug. when the ProxyOutputStream stream is already closed, calling write() will just cause a failure.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2214 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 14:11:10 +00:00
kohsuke
e83b388e6d
improved the error message.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2213 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 14:10:21 +00:00
kohsuke
7b8e851e31
fixed a test.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2212 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-15 13:55:10 +00:00
kohsuke
8c0a56f4ae
fixed a bug in the remoting support where the proxy returned from export() didn't work correctly if it uses user-defined classes as parameter values and return types.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2202 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-13 14:47:21 +00:00
kohsuke
862a646c7d
bug fix in the remoting mechanism.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@2126 71c3de6d-444a-0410-be80-ed276b4c234a
2007-02-09 20:28:06 +00:00
kohsuke
fe21d5c141
fixed a horrible horrible memory leak.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1807 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-19 19:27:34 +00:00
kohsuke
2f66019cf7
trying to fix a bug where the slave agent doesn't terminate itself even when the connection is lost.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1765 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-15 17:46:46 +00:00
ramapulavarthi
2a5e9bcce6
Fixed a bug where the wrong classloader was used to de-serialize the response object.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1735 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-12 02:43:24 +00:00
ramapulavarthi
98e7aa7d7f
Fixed a bug in handling exception thrown from UserRequest. The exception could be an instance of the user class, so it needs to be sent over on UserResponse, not the underlying Response.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1734 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-12 02:35:32 +00:00
kohsuke
fbc918b1f5
doc improvement.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1711 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-08 17:18:16 +00:00
kohsuke
f391d39411
added a test for resource support.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1599 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-03 05:57:45 +00:00
kohsuke
31ff889357
implemented resource support.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1598 71c3de6d-444a-0410-be80-ed276b4c234a
2007-01-03 05:57:34 +00:00
kohsuke
19860fa070
merging remoting-integration branch
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1523 71c3de6d-444a-0410-be80-ed276b4c234a
2006-12-29 19:15:53 +00:00
kohsuke
4cabee9f18
typo fixes.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1396 71c3de6d-444a-0410-be80-ed276b4c234a
2006-12-20 19:20:31 +00:00
kohsuke
183e0e5344
implemented another channel for doing computation locally.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1379 71c3de6d-444a-0410-be80-ed276b4c234a
2006-12-19 14:53:48 +00:00
kohsuke
8ddb883ead
doc improvement.
...
git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@1375 71c3de6d-444a-0410-be80-ed276b4c234a
2006-12-18 17:59:06 +00:00