Upgrade, so that indexing isn’t using as much CPU. That leaves more CPU for search.
Make sure you are on a recent release of Java. Run the G1 collector. If you need more throughput, add more replicas or use instance with more CPUs. Has the index gotten bigger since the move? wunder Walter Underwood wun...@wunderwood.org http://observer.wunderwood.org/ (my blog) > On Oct 28, 2018, at 8:21 PM, Parag Shah <parags.li...@gmail.com> wrote: > > The original question though is about performance issue in the Searcher. > How would you improve that? > > On Sun, Oct 28, 2018 at 4:37 PM Walter Underwood <wun...@wunderwood.org> > wrote: > >> The original question is for a three-node Solr Cloud cluster with >> continuous updates. >> Optimize in this configuration won’t help, it will just cause expensive >> merges later. >> >> I would recommend updating from Solr 4.4. that is a very early release for >> Solr Cloud. We saw dramatic speedups in indexing with 6.x. In early >> releases, the >> replicas actually did more indexing work than the leader. >> >> wunder >> Walter Underwood >> wun...@wunderwood.org >> http://observer.wunderwood.org/ (my blog) >> >>> On Oct 28, 2018, at 2:13 PM, Erick Erickson <erickerick...@gmail.com> >> wrote: >>> >>> Well, if you optimize on the master you'll inevitably copy the entire >>> index to each of the slaves. Consuming that much network bandwidth can >>> be A Bad Thing. >>> >>> Here's the background for Walter's comment: >>> >> https://lucidworks.com/2017/10/13/segment-merging-deleted-documents-optimize-may-bad/ >>> >>> Solr 7.5 is much better about this: >>> >> https://lucidworks.com/2018/06/20/solr-and-optimizing-your-index-take-ii/ >>> >>> Even with the improvements in Solr 7.5, optimize is still a very >>> expensive operation and unless you've measured and can _prove_ it's >>> beneficial enough to be worth the cost you should avoid it. >>> >>> Best, >>> Erick >>> On Sun, Oct 28, 2018 at 1:51 PM Parag Shah <parags.li...@gmail.com> >> wrote: >>>> >>>> What would you do if your performance is degrading? >>>> >>>> I am not suggesting doing this for a serving index. Only one at the >> Master, >>>> which ones optimized gets replicated. Am I missing something here? >>>> >>>> On Sun, Oct 28, 2018 at 11:05 AM Walter Underwood < >> wun...@wunderwood.org> >>>> wrote: >>>> >>>>> Do not run optimize (force merge) unless you really understand the >>>>> downside. >>>>> >>>>> If you are continually adding and deleting documents, you really do not >>>>> want >>>>> to run optimize. >>>>> >>>>> wunder >>>>> Walter Underwood >>>>> wun...@wunderwood.org >>>>> http://observer.wunderwood.org/ (my blog) >>>>> >>>>>> On Oct 28, 2018, at 9:24 AM, Parag Shah <parags.li...@gmail.com> >> wrote: >>>>>> >>>>>> Hi Mugeesh, >>>>>> >>>>>> Have you tried optimizing indexes to see if performance improves? It >>>>> is >>>>>> well known that over time as indexing goes on lucene creates more >>>>> segments >>>>>> which will be searched over and hence take longer. Merging happens >>>>>> constantly but continuous indexing will still introduce smaller >> segments >>>>>> all the time. Have your tried running "optimize" periodically. Is it >>>>>> something that you can afford to run? If you have a Master-Slave setup >>>>> for >>>>>> Indexer v/s searchers, you can replicate on optimize in the Master, >>>>> thereby >>>>>> removing the optimize load on the searchers, but replicate to the >>>>> searcher >>>>>> periodically. That might help with reducing latency. Optimize merges >>>>>> segments and hence creates a more compact index that is faster to >> search. >>>>>> It may involve some higher latency temporarily right after the >>>>> replication, >>>>>> but will go away soon after in-memory caches are full. >>>>>> >>>>>> What is the search count/sec you are seeing? >>>>>> >>>>>> Regards >>>>>> Parag >>>>>> >>>>>> On Wed, Sep 26, 2018 at 2:02 AM Mugeesh Husain <muge...@gmail.com> >>>>> wrote: >>>>>> >>>>>>> Hi, >>>>>>> >>>>>>> We are running 3 node solr cloud(4.4) in our production >> infrastructure, >>>>> We >>>>>>> recently moved our SOLR server host softlayer to digital ocean server >>>>> with >>>>>>> same configuration as production. >>>>>>> >>>>>>> Now we are facing some slowness in the searcher when we index >> document, >>>>>>> when >>>>>>> we stop indexing then searches is fine, while adding document then it >>>>>>> become >>>>>>> slow. one of solr server we are indexing other 2 for searching the >>>>> request. >>>>>>> >>>>>>> >>>>>>> I am just wondering what was the reason searches become slow while >>>>> indexing >>>>>>> even we are using same configuration as we had in prod? >>>>>>> >>>>>>> at the time we are pushing 500 document at a time, this processing is >>>>>>> continuously running(adding & deleting) >>>>>>> >>>>>>> these are the indexing logs >>>>>>> >>>>>>> 65497339 [http-apr-8980-exec-45] INFO >>>>>>> org.apache.solr.update.processor.LogUpdateProcessor – [rn0] >>>>> webapp=/solr >>>>>>> path=/update >>>>>>> params={distrib.from= >>>>>>> >>>>> >> http://solrhost:8980/solr/rn0/&update.distrib=FROMLEADER&wt=javabin&version=2&update.chain=dedupe >>>>>>> } >>>>>>> {add=[E4751FCCE977BAC7 (1612655281518411776), 8E712AD1BE76AB63 >>>>>>> (1612655281527848960), 789AA5D0FB149A37 (1612655281538334720), >>>>>>> B4F3AA526506F6B7 (1612655281553014784), A9F29F556F6CD1C8 >>>>>>> (1612655281566646272), 8D15813305BF7417 (1612655281584472064), >>>>>>> DD13CFA12973E85B (1612655281596006400), 3C93BDBA5DFDE3B3 >>>>>>> (1612655281613832192), 96981A0785BFC9BF (1612655281625366528), >>>>>>> D1E52788A466E484 (1612655281636900864)]} 0 9 >>>>>>> 65497459 [http-apr-8980-exec-22] INFO >>>>>>> org.apache.solr.update.processor.LogUpdateProcessor – [rn0] >>>>> webapp=/solr >>>>>>> path=/update >>>>>>> params={distrib.from= >>>>>>> >>>>> >> http://solrhost:8980/solr/rn0/&update.distrib=FROMLEADER&wt=javabin&version=2&update.chain=dedupe >>>>>>> } >>>>>>> {add=[D8AA2E196967D241 (1612655281649483776), E73420772E3235B7 >>>>>>> (1612655281666260992), DFDCF1F8325A3EF6 (1612655281680941056), >>>>>>> 1B10EF90E7C3695F (1612655281689329664), 51CBD7F59644A718 >>>>>>> (1612655281699815424), 1D31EF403AF13E04 (1612655281714495488), >>>>>>> 68E1DC3A614B7269 (1612655281723932672), F9BF6A3CF89D74FB >>>>>>> (1612655281737564160), 419E017E1F360EB6 (1612655281749098496), >>>>>>> 50EF977E5E873065 (1612655281759584256)]} 0 9 >>>>>>> 65497572 [http-apr-8980-exec-40] INFO >>>>>>> org.apache.solr.update.processor.LogUpdateProcessor – [rn0] >>>>> webapp=/solr >>>>>>> path=/update >>>>>>> params={distrib.from= >>>>>>> >>>>> >> http://solrhost:8980/solr/rn0/&update.distrib=FROMLEADER&wt=javabin&version=2&update.chain=dedupe >>>>>>> } >>>>>>> {add=[B63AD0671A5E57B9 (1612655281772167168), 00B8A4CCFABFA1AC >>>>>>> (1612655281784750080), 9C89A1516C9166E6 (1612655281798381568), >>>>>>> 9322E17ECEAADE66 (1612655281803624448), C6DDB4BF8E94DE6B >>>>>>> (1612655281814110208), DAA49178A5E74285 (1612655281830887424), >>>>>>> 829C2AE38A3E78E4 (1612655281845567488), 4C7B19756D8E4208 >>>>>>> (1612655281859198976), BE0F7354DC30164C (1612655281869684736), >>>>>>> 59C4A764BB50B13B (1612655281880170496)]} 0 9 >>>>>>> 65497724 [http-apr-8980-exec-31] INFO >>>>>>> org.apache.solr.update.processor.LogUpdateProcessor – [rn0] >>>>> webapp=/solr >>>>>>> path=/update >>>>>>> params={distrib.from= >>>>>>> >>>>> >> http://solrhost:8980/solr/rn0/&update.distrib=FROMLEADER&wt=javabin&version=2&update.chain=dedupe >>>>>>> } >>>>>>> {add=[1F694F99367D7CE1 (1612655281895899136), 2AEAAF67A6893ABE >>>>>>> (1612655281911627776), 81E72DC36C7A9EBC (1612655281926307840), >>>>>>> AA71BD9B23548E6D (1612655281939939328), 359E8C4C6EC72AFA >>>>>>> (1612655281954619392), 7FEB6C65A3E23311 (1612655281972445184), >>>>>>> 9B5ED0BE7AFDD1D0 (1612655281991319552), 99FE8958F6ED8B91 >>>>>>> (1612655282009145344), 2BDC61DC4038E19F (1612655282023825408), >>>>>>> 5131AEC4B87FBFE9 (1612655282037456896)]} 0 10 >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Sent from: http://lucene.472066.n3.nabble.com/Solr-User-f472068.html >>>>>>> >>>>> >>>>> >> >>
