Hi Pat

Thanks for the explanation. I see it's an issue with searchd being a
daemon (whoops!).

I know a bare minimum about unix, but I assume it's being signalled by
indexer & doesn't respond to it (so we don't know when it finishes).

However! Perhaps we could look for the log to be updated as in:

[Sat May 22 02:35:41.030 2010] [68934] rotating indices (seamless=1)
[Sat May 22 02:35:41.042 2010] [68934] rotating index 'volume_core':
success
[Sat May 22 02:35:41.048 2010] [68934] rotating index 'review_core':
success
[Sat May 22 02:35:41.054 2010] [68934] rotating index 'author_core':
success
[Sat May 22 02:35:41.060 2010] [68934] rotating index 'article_core':
success
[Sat May 22 02:35:41.060 2010] [68934] rotating finished

Also Aslak added a timeout trick to cucumber recently (0.7.3), could
be useful here in cases where something goes awry.
http://github.com/aslakhellesoy/cucumber/commit/3dd8613930482c63875dd5f4a8fcef0a47a77d65

What do you think?




On May 17, 9:25 am, Pat Allan <[email protected]> wrote:
> Hi Nick
>
> Unfortunately, it's not so much waiting for Sphinx to finish indexing - 
> that's a blocking request anyway - it's more waiting for searchd to load in 
> the index changes. Hence why it's a bit of guesswork.
>
> I know it's not ideal, but I've not thought of a better way around it yet. 
> I'd love suggestions, though :)
>
> --
> Pat
>
> On 15/05/2010, at 8:05 AM, nruth wrote:
>
>
>
>
>
> > Hi all
>
> > Many thanks to Pat et al for the pointers getting TS going with
> > cucumber, didn't take very long at all. I had some pre-TS scenarios
> > which were disabled, so added the gubbins and started using them
> > again, instant success.
>
> > I wonder about the sleep statements though. My computer seems to be
> > getting pretty medium-rare these days and I'm getting intermittently
> > failing scenarios from the indexing - failures seem to crop up when
> > the wind changes direction. I've increased the sleep period to
> > compensate, but it's a fairly arbitrary number & increasing it across
> > the board is wasteful over many executions of the feature set.
>
> > A better solution would be if ThinkingSphinx::Test.index could return
> > when sphinx has finished indexing, so the test waits for indexing to
> > finish before running the next steps.
>
> > Is this feasible, and patchable, or is there something in the way I'm
> > not aware of?
>
> > Regards
>
> > Nick
>
> > --
> > You received this message because you are subscribed to the Google Groups 
> > "Thinking Sphinx" group.
> > To post to this group, send email to [email protected].
> > To unsubscribe from this group, send email to 
> > [email protected].
> > For more options, visit this group 
> > athttp://groups.google.com/group/thinking-sphinx?hl=en.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Thinking Sphinx" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to 
> [email protected].
> For more options, visit this group 
> athttp://groups.google.com/group/thinking-sphinx?hl=en.

-- 
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/thinking-sphinx?hl=en.

Reply via email to