Connection resets about moji HOT 20 CLOSED

that's good to know, I will test it on Linux soon.

from moji.

I've never encountered this personally, but I've seen some reports in the last years that turned out to be issues with a firewall or other connection issues.
Your report is interesting because you tested it against the official client, though.

Can you post the stack trace, please?
Is this an especially large source code directory you're trying to upload? How long does it take before you receive an exception?

from moji.

I think I am observing the same issue, when I am trying to check around 800 source files. It takes around 4 seconds till I receive the exception.

The stack trace:
java.net.SocketException: Connection reset
at java.net.SocketInputStream.read(SocketInputStream.java:209)
at java.net.SocketInputStream.read(SocketInputStream.java:141)
at sun.nio.cs.StreamDecoder.readBytes(StreamDecoder.java:284)
at sun.nio.cs.StreamDecoder.implRead(StreamDecoder.java:326)
at sun.nio.cs.StreamDecoder.read(StreamDecoder.java:178)
at java.io.InputStreamReader.read(InputStreamReader.java:184)
at java.io.BufferedReader.fill(BufferedReader.java:161)
at java.io.BufferedReader.readLine(BufferedReader.java:324)
at java.io.BufferedReader.readLine(BufferedReader.java:389)
at moss.SocketClient.readFromServer(SocketClient.java:219)
at moss.SocketClient.sendQuery(SocketClient.java:381)

from moji.

Is there any resolution for this issue? Even I am facing the same issue of connection reset.

from moji.

mvn test - Tests run: 15, Failures: 11, Errors: 0, Skipped: 0
All are with "Connection reset"

from moji.

@nadman3
I am now trying to hunt this down in the "dev" branch. Please test this one as well.

This might also be an encoding or filesystem related problem that I cannot reproduce because of different JVM defaults. What's your JVM default encoding?
Can you attach the complete console output of "mvn test", please?

from moji.

https://docs.google.com/document/d/1w7-aRw91EWLfb0ckln29PhPsIbtPMK7EIh27sDOrnBA/edit?usp=sharing
Here is file with log.

And my JVM default encoding is UTF-8

from moji.

Thanks. That's a bug in the test code itself which is now fixed on the dev branch. It cannot have caused connection resets when running against the actual Moss server though.

This ticket will remain open until there is a fix for the sporadic connection resets that were reported, if there is any.
If you come across this problem, please try it again using the dev branch, report the issue here and attach

• the console output of the client
• the source code file that the client tried to upload last
• the "mvn test" console output

from moji.

• Client + "mvn test" output
• Source code files(zip)
• Client source code

from moji.

I cannot reproduce the exception with these files, neither on the dev branch nor on master. Didn't test it with your client though.
Do you call MOJI from your Swing draw thread or anything?

I am using the quickstart client with these settings:

Collection files = FileUtils.listFiles(new File(
"D:\temp\mossdevtest"), new String[] { "cpp" }, true);
socketClient.setLanguage("cc");

and one source code file per subdirectory.

from moji.

I noticed that on my Windows operating system all works well, but Linux does not. But on my friends Windows OS is the same error.
I also use these settings for quickstart

Collection files = FileUtils.listFiles(new File("D:\temp\mossdevtest"), new String[] { "cpp" }, true);
socketClient.setLanguage("cc");

And my directories looks like you said too...

and one source code file per subdirectory.

from moji.

But mvn test on Linux works without exceptions. Try to use your client on Linux with my source codes.

from moji.

I finally got around to try it on Linux. It worked with my test files, but I will try your test files tomorrow.

from moji.

Bug confirmed. Need to get a deeper look at the packet data to fix this.

from moji.

Might be fixed with the latest changes in the dev branch. I have to try some more test files to be sure.

@nadman3 : may I anonymize your test set and add it as a test?

from moji.

Of course you can. No problem. I'll try to use your version from dev branch

from moji.

Can you send me your test files? I'll test my program with them while program will not work correctly.

from moji.

v1.0.2 should fix this issue since it is binary compatible now, or so it seems.

I used the test set you gave me and compared the binary output for a dozen of files against the official Moss client.
And then I used the one from https://github.com/nordicway/SourceCode-Plagiarism-TestSets - but that always worked.

from moji.

Great! It really works. Thank you so much 👍
I think that issue may be closed

from moji.

You're welcome. Thanks for tracking it down with me!

from moji.