From: Mara K. <mar...@va...> - 2014-07-08 18:10:28
|
Sorry to say that we haven't made any progress on this. Any ideas where to start? On Jul 8, 2014 2:08 PM, "Stephen Ficklin" <spf...@gm...> wrote: > Hi Mara, > > Just curious if you were able to fix this issue? > > Stephen > > On 6/18/2014 7:09 PM, Mara Kim wrote: > > Hmm, that helped a bit, but it is still slow. > > This is especially surprising because an EXPLAIN ANALYZE of the > underlying query is very fast: > drupal=# EXPLAIN ANALYZE SELECT feature.uniquename AS feature_uniquename, > feature.name AS feature_name, feature.seqlen AS feature_seqlen, > feature.is_obsolete AS feature_is_obsolete > FROM chado.feature feature ORDER BY feature_name ASC LIMIT 25 OFFSET 0; > QUERY > PLAN > > --------------------------------------------------------------------------------------------------------------------------------------------------- > Limit (cost=0.00..11.54 rows=25 width=42) (actual time=0.099..0.248 > rows=25 loops=1) > -> Index Scan using feature_name_ind1 on feature > (cost=0.00..125462958.86 rows=271696608 width=42) (actual > time=0.097..0.242 rows=25 loops=1) > Total runtime: 0.312 ms > (3 rows) > > > > Whereas Drupal takes substantially longer with the same query: > *Query* > > SELECT feature.uniquename AS feature_uniquename, feature.name AS feature_name, feature.seqlen AS feature_seqlen, feature.is_obsolete AS feature_is_obsolete > FROM > {chado.feature} feature > ORDER BY feature_name ASC > LIMIT 25 OFFSET 0 > > *Title* Features *Path* This display has no path. *Query build time* 16.01 > ms *Query execute time* 46355.22 ms *View render time* 129.96 ms > > It seems that the Drupal query is using a different query plan...it may > not even be using the index... > > > > On Wed, Jun 18, 2014 at 2:00 PM, Stephen Ficklin <spf...@gm...> > wrote: > >> Oh, right. I remember you had this problem previously. The problem is >> that the view which provides the search page, by default, wants to provide >> a feature type drop down with all of the feature types that you have >> available. So, it spends a lot of time going through all of your features >> to find all of the types which slows the view down. Most likely you don't >> want to search by all your feature types anyway, so I think you can speed >> it up by the following: >> >> 1) Click the little wheel that appears in the top-left corner of the form >> when you mouse over it and click 'Edit view'. This will take you to the >> View Edit page. >> 2) Under the filter criteria section click the 'Chado Feature: Type Id >> (exposed)' link. >> 3) make the following changes: Select the radio button 'Grouped >> filters', then scroll a bit and in the table add in the feature types you >> want to allow the user to search by and select the proper value in the drop >> down. Click the apply button and the overlay will close. >> 4) Finally, click the 'Save' button on the top right hand corner. >> >> This should take you back to the search page which hopefully will load >> faster. >> >> Stephen >> >> >> On 6/18/2014 2:43 PM, Mara Kim wrote: >> >> Hmm, so the Sequence Features search does find the genes. However, it >> loads *extremely* slowly. Is there a way to optimize the load time of this >> page? >> >> >> On Tue, Jun 17, 2014 at 1:41 PM, Stephen Ficklin <spf...@gm...> >> wrote: >> >>> Hi Mara, >>> >>> What about finding features via the Sequence Features search page? It >>> appears under the 'Search Data' menu, or at http://[your >>> site]/chado/feature >>> >>> Stephen >>> >>> >>> On 6/17/2014 2:25 PM, Mara Kim wrote: >>> >>> It would be preferable for this to be fixed as soon as possible since >>> finding features using the Feature Browser is a bit unwieldy. However, >>> this could be alleviated in the interim by adding alphabetical links in the >>> feature browser so that it is easier to find specific features. >>> >>> >>> On Tue, Jun 17, 2014 at 8:06 AM, Stephen Ficklin <spf...@gm...> >>> wrote: >>> >>>> Hi Mara, >>>> >>>> Thanks for cluing us in. I have added this bug to the Tripal "Issues" >>>> tracker on the Drupal website. You can subscribe to the page to see any >>>> updates when this bug is addressed: https://drupal.org/node/2287615. >>>> If >>>> it's a critical issue to you let me know and we can try to get to it >>>> quicker. >>>> >>>> Stephen >>>> >>>> On 6/11/2014 1:16 PM, Mara Kim wrote: >>>> > Hello, >>>> > >>>> > I believe there is another bug on the 7.x-2.x branch, commit >>>> > a68b9af2f768d244ddbc7607dd94d31de55673b2. >>>> > >>>> > The drupal cron job that adds new pages to the search index no longer >>>> adds >>>> > additional nodes. This definitely used to work in an older commit, >>>> > since I can find pages that Tripal created in the past via the drupal >>>> > search tool. Running cron.php manually via >>>> > Administration->Configuration->System->Cron completes successfully, >>>> > but does not add the pages to the search index. >>>> > >>>> > Examination of the reflog indicates that this bug was introduced >>>> > sometime after commit 0aa9fbf9f197f92363b3adc887e36325e6930a9d >>>> > >>>> > Thanks, >>>> > Mara Kim >>>> > >>>> > Ph.D. Candidate >>>> > Computational Biology >>>> > Vanderbilt University >>>> > Nashville, TN >>>> > >>>> > >>>> > >>>> ------------------------------------------------------------------------------ >>>> > HPCC Systems Open Source Big Data Platform from LexisNexis Risk >>>> Solutions >>>> > Find What Matters Most in Your Big Data with HPCC Systems >>>> > Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. >>>> > Leverages Graph Analysis for Fast Processing & Easy Data Exploration >>>> > http://p.sf.net/sfu/hpccsystems >>>> > _______________________________________________ >>>> > Gmod-tripal-devel mailing list >>>> > Gmo...@li... >>>> > https://lists.sourceforge.net/lists/listinfo/gmod-tripal-devel >>>> >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> HPCC Systems Open Source Big Data Platform from LexisNexis Risk >>>> Solutions >>>> Find What Matters Most in Your Big Data with HPCC Systems >>>> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. >>>> Leverages Graph Analysis for Fast Processing & Easy Data Exploration >>>> http://p.sf.net/sfu/hpccsystems >>>> _______________________________________________ >>>> Gmod-tripal-devel mailing list >>>> Gmo...@li... >>>> https://lists.sourceforge.net/lists/listinfo/gmod-tripal-devel >>>> >>>> >>> >>> >>> -- >>> Mara Kim >>> >>> Ph.D. Candidate >>> Computational Biology >>> Vanderbilt University >>> Nashville, TN >>> >>> >>> >> >> >> -- >> Mara Kim >> >> Ph.D. Candidate >> Computational Biology >> Vanderbilt University >> Nashville, TN >> >> >> > > > -- > Mara Kim > > Ph.D. Candidate > Computational Biology > Vanderbilt University > Nashville, TN > > > |