ts:rebuild works great, but it takes too long with over 100,000 records. It should be possible to have a consistent index that is only rebuilt once per day.
On Oct 15, 11:38 am, Beau Frusetta <[email protected]> wrote: > Have you tried "ts:rebuild" instead of "ts:reindex"? > > On Oct 15, 2010, at 11:34 AM, claytonlz wrote: > > > > > Here's the output of that ls: > > > -rw-r--r-- 1 deploy deploy 171864 Oct 15 13:00 post_core.spa > > -rw-r--r-- 1 deploy deploy 234975 Oct 15 13:00 post_core.spd > > -rw-r--r-- 1 deploy deploy 571 Oct 15 13:00 post_core.sph > > -rw-r--r-- 1 deploy deploy 13421 Oct 15 13:00 post_core.spi > > -rw-r--r-- 1 deploy deploy 0 Oct 15 13:00 post_core.spk > > -rw------- 1 deploy deploy 0 Oct 15 13:00 post_core.spl > > -rw-r--r-- 1 deploy deploy 0 Oct 15 13:00 post_core.spm > > -rw-r--r-- 1 deploy deploy 273326 Oct 15 13:00 post_core.spp > > -rw-r--r-- 1 deploy deploy 0 Oct 15 13:00 post_delta.spa > > -rw-r--r-- 1 deploy deploy 1 Oct 15 13:00 post_delta.spd > > -rw-r--r-- 1 deploy deploy 571 Oct 15 13:00 post_delta.sph > > -rw-r--r-- 1 deploy deploy 1 Oct 15 13:00 post_delta.spi > > -rw-r--r-- 1 deploy deploy 0 Oct 15 13:00 post_delta.spk > > -rw------- 1 deploy deploy 0 Oct 15 13:00 post_delta.spl > > -rw-r--r-- 1 deploy deploy 0 Oct 15 13:00 post_delta.spm > > -rw-r--r-- 1 deploy deploy 1 Oct 15 13:00 post_delta.spp > > > We are storing the indexes in a shared path above the current release. > > That path is symlinked to the current release upon deploy. > > > We're basically following your deployment strategy, ours is something > > like: > > > ...regular cap deploy steps... > > > 1. deploy:cleanup > > 2. ts:conf > > 3. ts:reindex > > > So we're not stopping and starting sphinx before we symlink and > > configure. Could that be the problem? > > > Thanks, > > Clayton > > > On Oct 14, 5:20 pm, Pat Allan <[email protected]> wrote: > >> Hi Clayton > > >> I'm not quite sure what's causing this... as the error is pointing to the > >> core index, not the delta (my first guess was the regular delta updates > >> were conflicting with the full index - perhaps that's still the case, but > >> not so convinced at this point). > > >> What's the output of the following on your server: > > >> ls -alh /home/deploy/apps/my_apps/current/db/sphinx/production/ > > >> Also: I recommend storing production indexes outside of the releases > >> directory - have a read through of the following documentation: > > >> http://freelancing-god.github.com/ts/en/deployment.html > > >> Cheers > > >> -- > >> Pat > > >> On 14/10/2010, at 4:34 AM, claytonlz wrote: > > >>> I've got a sphinx installation using thinking sphinx that has ~100,000 > >>> records distributed across a number of models. One of my models is > >>> frequently updated causing there to be numerous delta indexes created. > >>> When ThinkingSphinx runs ts:reindex once per hour, it almost always > >>> results in this error for my frequently updated model: > > >>> WARNING: rotating index 'post_core': cur to old rename failed: rename / > >>> home/deploy/apps/my_app/releases/20101008210428/db/sphinx/production/ > >>> post_core.spl to /home/deploy/apps/my_app/releases/20101008210428/db/ > >>> sphinx/production/post_core.old.spl failed: No such file or directory > > >>> This error occurs again and again over the course of 30 minutes with > >>> multiple messages like this in the log file. It's only when I > >>> completely rebuild in the index that this problem goes away. > > >>> What can I do to prevent this from happening? As it is right now, > >>> things go smoothly for a while and then the indexes fail to rotate and > >>> the search functionality becomes useless because of all of the missing > >>> records. > > >>> -- > >>> 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.
