On 09/ 7/10 02:56 PM, [email protected] wrote:
Hi Shawn,
On Sat, Sep 04, 2010 at 11:20:21PM -0700, Shawn Walker wrote:
webrev:
http://cr.opensolaris.org/~swalker/pkg-16982/
This looks good. Everything that follows are just nits.
man/pkg.depotd.1m.txt:
...
server/repository.py:
Sorry if this is a dumb question, but are the changes in this file to
keep it consistent with the locking interface you've put in the
indexer? I'm still abosrbing my coffee, and I was trying to figure
out why this changed from blocking=True to blocking=False as the
default.
Because there's no timeout on blocking and indexing can take a long
time. In most cases, it's better to error out immediately then wait a
long time for a lock.
As such, my approach was that some operations I was willing to block on
the lock for (such as adding packages) but in general, I didn't think it
should.
server/depot.py:
- line 321: The code will log a message when the IndexLockedException
is raised. Would it make sense to log a message when Indexing can't be
started because the Queue is full too?
Yes, I've already changed this to simply log a message. I'm not
anticipating that we'll ever actually hit a queue full case for this as
it only ever gets called during depot startup which is why I wasn't
concerned about raising an error.
-Shawn
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
