Bazaar

  1. Bug #255292
  2. Comment #3

Comment 3 for bug 255292

Revision history for this message
John A Meinel (jameinel) wrote : Re: bzr crashes while pushing up a new revisionof get-you

Sure, I realize it is bidirectional communication. ...

I just thought of something, though. Could this be when triggering an auto-pack?

Even if it was, I would still be surprised for something like this to be transmitting >2GB in a single packet.

Also, I didn't think sending packets that large was even allowed for sftp (perhaps this is just raw ssh?) At least, what I remember of the sftp spec was that you couldn't guarantee readv > 32K (though openssh allowed 64k requests for readv)

I'm somewhat tempted to say bogus sftp server.

Also look closer at the traceback, and you can see that we are reading the lock/held/info file. I think what is happening is that we are peeking at a file that doesn't exist, and the SFTP server is sending back garbage, rather than a "ENOENT".

 File "/usr/lib/python2.5/site-packages/bzrlib/lockdir.py", line 395, in confirm
    info = self.peek()
  File "/usr/lib/python2.5/site-packages/bzrlib/lockdir.py", line 418, in peek
    info = self._read_info_file(self._held_info_path)
  File "/usr/lib/python2.5/site-packages/bzrlib/lockdir.py", line 408, in _read_info_file
    return self._parse_info(self.transport.get(path))

At a minimum, the 'info' file should be on the order of *maybe* 10kB. Certainly nowhere near 2GB in size.