Your message dated Tue, 28 Oct 2008 18:21:48 +0100
with message-id <[EMAIL PROTECTED]>
and subject line good and bad news about gvfs and Clint's NMU
has caused the Debian Bug report #502725,
regarding gvfs hangs for mounts after heavy use
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [EMAIL PROTECTED]
immediately.)
--
502725: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=502725
Debian Bug Tracking System
Contact [EMAIL PROTECTED] with problems
--- Begin Message ---
Package: gvfs
Version: 0.2.5-1
Severity: serious
Reasoning: cripples things that might wait for it
Hi,
when running the attached program with ~/.gvfs/localtest/mirror/debian (or
whereever you
have a sizable collection of files) for a while, gvfsd becomes unresponsive and
apparently
does not answer dbus requests for that gvfs-mount anymore. This happens with
multiple
backends (e.g. sftp and localtest). Other mounts appear to be still working.
This does not seem to be a mere gvfs-fuse-daemon-problem because gvfs-ls is
affected as
well. I noticed this problom while testing a patch for the gvfs-fuse RC bug.
Error message is
$ gvfs-ls localtest:/home
Error: Error while getting peer-to-peer dbus connection: Did not receive a
reply. Possible
causes include: the remote application did not send a reply, the message bus
security
policy blocked the reply, the reply timeout expired, or the network connection
was broken.
Kind regards
T.
--
Thomas Viehmann, http://thomas.viehmann.net/
#!/usr/bin/python
import os, sys, glob, random, thread
BLOCK = 1024*1024
def madseek(rootdir):
while 1:
p = rootdir
while os.path.isdir(p):
d = glob.glob(os.path.join(p, '*'))
if not d:
break
p = d[random.randint(0,len(d)-1)]
if os.path.isfile(p):
print p
size = os.stat(p).st_size
f = open(p)
pos = random.randint(0,max(size-BLOCK,0))
f.seek(pos)
f.read(min(BLOCK,size-pos))
f.close()
a = map(lambda x: thread.start_new_thread(madseek, (sys.argv[1],)), range(10))
madseek(sys.argv[1])
--- End Message ---
--- Begin Message ---
Version: 0.2.5-1.1
reopen 496269
notfixed 496269 0.2.5-1.1
found 496269 0.2.5-1.1
Hi,
the bug I reported (#502725) is indeed fixed with Clint's NMU, thanks. I
should note that gvfs-fuse-daemon still hangs on heavy use (i.e. #496269
is not fixed, demonstration: run my test script from #502725 on
~/.gvfs/sftp on localhost/path/to/debian-mirror for a while. The ~/.gvfs
will hang while ls-gvfs still works as expected). Note that testing with
localtest may be tricky: I have found that localtest crashes much faster
and that for those gvfs-ls then also stops working (hinting at a bug in
either gvfsd or the localtest backend, the latter evidence making the
backend more suspicious because accessing other mounts still works, but
that is no proof).
Kind regards
T.
--
Thomas Viehmann, http://thomas.viehmann.net/
--- End Message ---
