DuckDB shine tsarin sarrafa bayanai na SQL OLAP a cikin samarwa
An sanar da sakin sabon sigar DBMS DuckDB 0.6.0, sigar da an inganta matse bayanai, ban da gaskiyar cewa an ƙara sababbin ayyuka, da kuma inganta kayan ajiya, a tsakanin sauran abubuwa.
DuckDB hada SQLite Properties kamar ƙaƙƙarfan ƙarfi, ikon haɗawa a cikin nau'in ɗakin karatu mai haɗaka, adana bayanan bayanai a cikin fayil guda ɗaya da kuma ingantacciyar hanyar sadarwa ta CLI, tare da kayan aiki da haɓakawa don yin tambayoyin nazari wanda ke rufe wani muhimmin sashi na bayanan da aka adana, alal misali, wanda ke yin tara duk abin da ke cikin tebur ko haɗa manyan teburi masu yawa.
Babban sabbin fasalulluka na DuckDB 0.6.0
A cikin wannan sabon sigar da aka gabatar, an yi nuni da cewa aikin ya ci gaba da inganta tsarin ajiya, Bayan haka an aiwatar da yanayin rubuta faifai, inda a lokacin da aka loda babban saitin bayanai a cikin wata ma'amala guda ɗaya, ana matsa bayanan kuma ana watsa su zuwa fayil daga ma'ajin bayanai ba tare da jiran umarnin COMMIT don yin ciniki ba.
Wani sauye-sauyen da suka yi fice a sabuwar sigar ita ce ƙarin goyon baya don loda bayanai a layi daya cikin tebur daban, wanda zai iya ƙara haɓaka saurin saukewa akan tsarin multicore. Misali, a cikin tsohon sigar, loda rumbun adana bayanai da layuka miliyan 150 a kan CPU mai nauyin 10-core ya dauki dakika 91, kuma a cikin sabon sigar, wannan aiki yana daukar dakika 17. Akwai hanyoyi guda biyu na loda masu daidaitawa: tare da adana oda kuma ba tare da kiyaye oda ba.
Don matsawa bayanai, ana amfani da FSST algorithm (Static Symbols Quick Table), wanda ke ba ka damar tattara bayanai a cikin layuka ta amfani da ƙamus na gama gari na nau'in matches. Aikace-aikacen sabon algorithm ya ba da izinin rage girman bayanan gwajin daga 761 MB zuwa 251 MB.
Don damfara lambobi (DOUBLE da FLOAT) ana ba da shawarar Chimp da Patas algorithms. Idan aka kwatanta da algorithm na Gorillas na baya, Chimp yana samar da babban matakin matsawa da kuma saurin ragewa. Algorithm na Patas yana bayan Chimp cikin sharuddan matsawa, amma yana da matukar sauri cikin saurin raguwa, wanda kusan daidai yake da karanta bayanan da ba a matsawa ba.
Hakanan an lura cewa an ƙara ikon gwaji don loda bayanai daga fayilolin CSV a cikin rafukan layi ɗaya da yawa (SET experimental_parallel_csv=gaskiya), wanda ke rage lokacin lodi ga manyan fayilolin CSV. Misali, lokacin da zaɓin ya kunna, lokacin zazzagewar fayil ɗin CSV 720MB ya ragu daga daƙiƙa 3,5 zuwa daƙiƙa 0,6.
Na sauran canje-canje wanda ya fice daga wannan sabon sigar:
- An aiwatar da yuwuwar aiwatar da layi ɗaya na ƙirƙirar ƙididdiga da ayyukan gudanarwa.
- SQL yana ba da ikon samar da tambayoyin da suka fara da kalmar "DAGA" maimakon "Zabi". A wannan yanayin, ana tsammanin tambayar zata fara da "SELECT *".
- Ƙara goyon baya don kalmar "COLUMNS" a cikin SQL, yana ba ku damar yin aiki akan ginshiƙai da yawa ba tare da kwafin magana ba.
- Ingantaccen amfani da ƙwaƙwalwar ajiya. Ta hanyar tsoho akan dandamali na Linux, ana amfani da ɗakin karatu na jemalloc don sarrafa ƙwaƙwalwar ajiya. Ingantacciyar ingantacciyar aikin hash na ayyukan hash lokacin da ƙwaƙwalwar ajiya ta iyakance.
- An ƙara yanayin fitarwa na ".mode duckbox" zuwa CLI, watsar da ginshiƙan tsakiya dangane da faɗin layukan taga tasha). Tare da sigar ".maxrows X", za ku iya iyakance adadin layuka masu fitarwa.
- CLI tana ba da shigarwar atomatik-sane-dare (kalmomin maɓalli, sunayen tebur, ayyuka, sunayen shafi, da shigar da sunayen fayil an kammala).
- An kunna CLI ta tsohuwa don nuna alamar ci gaban tambaya.
a karshe idan kun kasance sha'awar ƙarin sani game da shi, zaka iya duba bayanan A cikin mahaɗin mai zuwa.