timeout for LockContention retry is always 0 for smart server
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Bazaar |
Won't Fix
|
Undecided
|
Unassigned |
Bug Description
The timeout for LockContention retry is always 0 for smart server. Due to this, the retry does not happen.
I put some checks in bzrlib.
if timeout is None:
timeout = _DEFAULT_
While lockdir.py initialized _DEFAULT_
I am using trunk (2.2b3) on Lucid.
Output for trunk
=============
[foo]% date
Thu Jun 10 17:15:28 IST 2010
[foo]% ~/src/bzr.
parthm@localhost's password:
Unable to obtain lock filtered-
held by Parth Malwankar <email address hidden> on host parthm-laptop [process #9606]
locked 26 hours, 27 minutes ago
Will continue to try until 17:15:33, unless you press Ctrl-C.
See "bzr help break-lock" for more.
bzr: ERROR: Could not acquire lock "(remote lock)":
[foo]% date
Thu Jun 10 17:15:34 IST 2010
[foo]% ssh parthm@localhost "which bzr"
parthm@localhost's password:
/home/parthm/
[foo]%
Output with debug info
==================
[foo]% which bzr
/home/parthm/
[foo]% ssh parthm@localhost "which bzr"
parthm@localhost's password:
/home/parthm/
[foo]% date
Thu Jun 10 17:23:40 IST 2010
[foo]% bzr push bzr+ssh:
parthm@localhost's password:
Unable to obtain lock held by Parth Malwankar <email address hidden> at parthm-laptop [process #9606], acquired 26 hours, 35 minutes ago.
Will continue to try until 17:23:51#current time: 17:23:51#timeout: 0, unless you press Ctrl-C.
See "bzr help break-lock" for more.
bzr: ERROR: Could not acquire lock "(remote lock)": bzr+ssh:
[foo]% date
Thu Jun 10 17:23:54 IST 2010
[foo]%
Note that in the "Will continue.." line, "try until" is the same as "current time" as timeout is 0.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Parth Malwankar wrote: lockdir. wait_lock. TIMEOUT_ SECONDS TIMEOUT_ SECONDS to 300, the actual value at runtime is 0.
> Public bug reported:
>
> The timeout for LockContention retry is always 0 for smart server. Due to this, the retry does not happen.
> I put some checks in bzrlib.
> if timeout is None:
> timeout = _DEFAULT_
> While lockdir.py initialized _DEFAULT_
>
> I am using trunk (2.2b3) on Lucid.
>
This is actually by design. The problem is that it was hanging on the
server side, without having the user know what they could do.
If you look closely, the default is overridden inside 'cmd_serve'.
status: wontfix
John
=:->
-----BEGIN PGP SIGNATURE----- enigmail. mozdev. org/
Q9VoACgkQJdeBCY SNAAP5UQCg1Lz7/ BR8p01UqXUs1bLM jk7H 9RnNkwCuNtfO/ 2S4L
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://
iEYEARECAAYFAkw
EioAn3HwMscsPL4
=WT3D
-----END PGP SIGNATURE-----