Re: rsync and pbulk

[Richard PALO <richard%netbsd.org@localhost>]

Le 23/10/15 17:36, Joerg Sonnenberger a écrit :
> On Fri, Oct 23, 2015 at 03:29:51PM +0200, Richard PALO wrote:
>> I actually had better success with the pure joyent patches.
>> I applied both the earlier and this mornings patch that joerg did:
> 
> That looks a build that does nothing. I never got around to shuffle the
> init logic around a bit to check for jobs before running the start
> script. Essentially, the first client would check for a job, finds
> nothing and triggers the shutdown. That's why the others are getting a
> connection refused.
> 
> Joerg
> 

In the joyent patches, prior to integrating your recent patches, I could always
rebuild a package that failed initially and usually be able to restart... 
more or less where I left off 
(I had some problemes initially but I believe at the time I was not
on the most current joyent branch and not updating pbulk-base only pbulk).

With the current errors, I have no choice but to start all over, in my current
case (bulk-medium + my commonly used packages) that's hours and hours lost.

It would naturally be nice to be able to do both after a build with any failures.
That is, both bulkbuild-rebuild *and* bulkbuild-restart.

Also, although it may be more related to joyents patches than otherwise, it would
be nice if for any reason after invoking a bulkbuild script and subsequently aborting
e.g. ^C...  upon the next invocation, the chroots are in the way needing cleanup -
currently manually, which when many is a nuissance.

cheers
-- 
Richard PALO