bazaar locks...

[Michael Andronov]

Thanks for a speedy reply.

The 'bzr push ...'  is called from within the bash shell.
Is there simple way to check the return status of the  'bzr push..'?
Should I parse err. stream?
Or should I re-do the whole script and use Python bzrlib to execute
the above command, and then check for the return?

I'm wondering what is more robust/common/easier?

Thanks.

Michael.


On Wed, Sep 8, 2010 at 12:55 PM, Jelmer Vernooij <jelmer at samba.org> wrote:
> On Wed, 2010-09-08 at 12:41 -0400, Michael Andronov wrote:
>> I'm periodically observing the following issue while pushing to remote server:
>> "
>> ...
>> bzr: ERROR: Could not acquire lock "LockDir(sftp://michaela@....
>> /.bzr/branch/lock)":
>> Unable to obtain lock
>> sftp://michaela@localhost.../.../.../.../.../.bzr/branch/lock
>> held by michaela at Michael-Andronovs-Mac-mini.local on host
>> Michael-Andronovs-Mac-mini.local [process #60568]
>> locked 145 hours, 59 minutes ago
>> Will continue to try until 11:47:33, unless you press Ctrl-C.
>> See "bzr help break-lock" for more.
>> ...
>> "
>> I'm wondering  -
>> a. What is the most common problem, leading to that? Broken process?
> A process that's pushing to the branch died without unlocking the branch
> properly. Examples of how this can happen include sudden network
> disconnects, client getting killed without having time to clean up
> (segfault, power switch turned off, kill -9, etc).
>
>> b. What would be  the most common algorithm for recovery? ( assuming
>> that the 'pushes' are done automatically by computer).
> run "bzr break-lock <url>".
>
> Cheers,
>
> Jelmer
> b
>