Ukukhishwa kwe-DBMS , edizayinelwe ukugcina nokucubungula idatha ngendlela yochungechunge lwesikhathi (izingcezu zamanani epharamitha ngezikhathi ezithile; isikhathi sakha irekhodi kanye nesethi yamanani ahambisana nalesi sikhathi). Lolu hlobo lwesitoreji lulungele izinhlelo zokusebenza ezifana nezinhlelo zokuqapha, izinkundla zokuhweba, amasistimu okuqoqa amamethrikhi nezinzwa. Amathuluzi okuhlanganiswa nephrojekthi anikeziwe и .
Iphrojekthi ye-TimescaleDB isetshenziswa njengesandiso ku-PostgreSQL kanye ilayisensi ngaphansi kwe-Apache 2.0. enezici ezithuthukisiwe ezitholakala ngaphansi kwelayisensi yobunikazi ehlukile (I-TSL), engavumeli izinguquko, ivimbela ukusetshenziswa kwekhodi emikhiqizweni yezinkampani zangaphandle futhi ayikuvumeli ukusetshenziswa kwamahhala kusizindalwazi samafu (database-as-a-service).
Phakathi kwezinguquko ku-TimescaleDB 1.7:
- Ukwesekwa okwengeziwe kokuhlanganiswa ne-DBMS . Ukusekela kwe-PostgreSQL 9.6.x kanye ne-10.x kuhoxisiwe (I-Timescale 2.0 izosekela i-PostgreSQL 11+ kuphela).
- Ukuziphatha kwemibuzo eqhuba imisebenzi ehlanganisiwe eqhubekayo (ukuhlanganiswa kwedatha engenayo ngokuqhubekayo ngesikhathi sangempela) kushintshiwe. Imibuzo enjalo manje ihlanganisa ukubukwa okuphathekayo nedatha esanda kufika engakenziwa (ngaphambilini, ukuhlanganisa bekumboze kuphela idatha eseyivele yenziwe). Ukuziphatha okusha kusebenza ekuhlanganiseni okusha okudaliwe okuqhubekayo; ekubukeni okukhona, ipharamitha ethi “timescaledb.materialized_only=false” kufanele isethwe ngokuthi “ALTER VIEW”.
- Amanye amathuluzi okuphatha umjikelezo wokuphila wedatha athuthukisiwe adluliselwe kunguqulo Yomphakathi esuka kunguqulo yezentengiso, okuhlanganisa ikhono lokuhlanganisa kabusha idatha nokucubungula izinqubomgomo zokukhipha idatha eziphelelwe yisikhathi (okukuvumela ukuthi ugcine idatha yamanje kuphela futhi ususe ngokuzenzakalelayo, uhlanganise noma ubeke kungobo yomlando amarekhodi aphelelwe yisikhathi).
Masikhumbule ukuthi i-TimescaleDB DBMS ikuvumela ukuthi usebenzise imibuzo egcwele ye-SQL ukuze uhlaziye idatha eqoqiwe, uhlanganise kalula ukusetshenziswa okukhona kuma-DBMS ahlobene nokukala namandla atholakala ezinhlelweni ezikhethekile ze-NoSQL. Isakhiwo sesitoreji sithuthukisiwe ukuze kuqinisekiswe isivinini esikhulu sokwengezwa kwedatha. Isekela ukungezwa kweqoqo lamasethi edatha, ukusetshenziswa kwezinkomba zememori, ukulayishwa kabusha kwezingcezu zomlando, kanye nokusetshenziswa kokwenziwe.
Isici esibalulekile se-TimescaleDB ukusekela kwayo ukwahlukanisa okuzenzakalelayo kohlu lwedatha. Ukusakaza kwedatha okokufaka kusatshalaliswa ngokuzenzakalela kuwo wonke amathebula ahlukanisiwe. Izigaba zidalwe ngokuya ngesikhathi (isigaba ngasinye sigcina idatha yesikhathi esithile) noma ngokuhlobene nokhiye othize (ngokwesibonelo, i-ID yedivayisi, indawo, njll.). Ukuze kuthuthukiswe ukusebenza kahle, amathebula ahlukanisiwe angasatshalaliswa kumadiski ahlukene.
Ngemibuzo, isizindalwazi esihlukanisiwe sibukeka njengetafula elilodwa elikhulu elibizwa ngokuthi i-hypertable. I-hypertable iwumfanekiso obonakalayo wamathebula ngamanye aqongelela idatha engenayo. I-hypertable ayisetshenziselwa kuphela imibuzo nokwengeza idatha, kodwa futhi nemisebenzi efana nokudala izinkomba nokushintsha isakhiwo (“ALTER TABLE”), kufihlwa isakhiwo esinezigaba eziphansi sesizindalwazi kunjiniyela. Nge-hypertable, ungasebenzisa noma yimiphi imisebenzi ehlanganisiwe, imibuzo engaphansi, ukuhlanganisa imisebenzi (JOIN) namathebula avamile, nemisebenzi yewindi.
Source: opennet.ru