This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.

Bug 196453 - Can not build remote project: protocol error
Summary: Can not build remote project: protocol error
Status: RESOLVED FIXED
Alias: None
Product: cnd
Classification: Unclassified
Component: Remote (show other bugs)
Version: 7.0
Hardware: PC Windows XP
: P3 normal (vote)
Assignee: Vladimir Kvashin
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-03-08 20:34 UTC by Vladimir Kvashin
Modified: 2011-07-26 13:48 UTC (History)
1 user (show)

See Also:
Issue Type: DEFECT
Exception Reporter:

Attachments
The proposed fix (4.16 KB, patch)
2011-06-05 23:53 UTC, Vladimir Kvashin 		Details | Diff
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Vladimir Kvashin 2011-03-08 20:34:02 UTC 
While testing NetBeans on Windows XP (in virtual box), building remotely on Ubuntu (host system in VBox terms), I've got the following issue quite stable:

Process exited unexpectedly while file info was being sent
protocol error: first line shoud start with VERSION: D /home/vk155xxx/.netbeans/remote/71.15.28.29/hostname-Windows-x86/C/Documents and Settings

Initialization error
 BUILD FAILED
Comment 1 Vladimir Kvashin 2011-03-08 20:36:46 UTC 
I've just remembered that there already was an issue 187439 with the same error message. But it seems its root cause was found and fixed.
Comment 2 Vladimir Kvashin 2011-03-08 21:25:17 UTC 
For some reason remote process (rfs_controller) isn't in time to read the first line of input that is send to it.

The process is created (NativeProcessBuilder.call() is onvoked) in the same thread in which input is sent to it. When i tried to get pid and state of the process just before sending input, pid already existed and state was "running".
Comment 3 Vladimir Kvashin 2011-03-08 21:45:38 UTC 
I've just pushed it  a workaround - a 200 ms sleep
http://hg.netbeans.org/cnd-main/rev/f52df47afd3a

With it I wasn't able to get build fail.

But what I'm afraid of is that there is a race somewhere in 
nativeexecution or jsch. I'm not closing the issue - we'd better find and fix root cause than work it around. The question is whether we'll be in time to do this in 7.0

To switch workaround off, use
-J-Drfs.instable.sleep=0
Comment 4 Quality Engineering 2011-03-10 09:34:19 UTC 
Integrated into 'main-golden', will be available in build *201103100400* on http://bits.netbeans.org/dev/nightly/ (upload may still be in progress)
Changeset: http://hg.netbeans.org/main/rev/f52df47afd3a
User: Vladimir Kvashin <vkvashin@netbeans.org>
Log: A workaround for #196453 - Can not build remote project: protocol error
Comment 5 Leonid Lenyashin 2011-03-12 12:51:08 UTC 
Is it a HR fix candidate? Do you expect a fix or workaround is enough for now?
Comment 6 Vladimir Kvashin 2011-03-14 08:40:06 UTC 
Yes, I believe it's a HR candidate.
The workaround seems to work, so for now its enough I believe.
Comment 7 Vladimir Kvashin 2011-03-15 14:01:15 UTC 
After a discussion with QA it was decided to downgrade to P3 as it seems to reproduce in a very specific configuration.

However I believe it should be fixed in 7.01 or patch if we have one of these.
Comment 8 Vladimir Kvashin 2011-06-05 23:53:32 UTC 
Created attachment 108741 [details]
The proposed fix
Comment 9 Quality Engineering 2011-06-06 04:50:06 UTC 
Integrated into 'releases'
Changeset: http://hg.netbeans.org/releases/rev/
User: 
Log:
Comment 11 Vladimir Kvashin 2011-06-06 08:15:15 UTC 
http://hg.netbeans.org/releases/rev/46b7b7707fd9
Comment 12 Quality Engineering 2011-06-07 03:01:05 UTC 
Integrated into 'main-golden'
Changeset: http://hg.netbeans.org/main-golden/rev/
User: 
Log: