Ka otu afọ nke mmepe gachara, e bipụtala alaka ọhụrụ kwụsiri ike nke PostgreSQL 15 DBMS. A ga-ewepụta mmelite maka alaka ọhụrụ ahụ n'ime afọ ise ruo Nọvemba 2027.
Isi ihe ọhụrụ:
- Nkwado agbakwunyere maka iwu SQL "MERGE", nke yiri okwu "INSERT ... NA Ọgbaghara". MERGE na-enye gị ohere ịmepụta nkwupụta SQL nwere ọnọdụ na-ejikọta INSERT, emelitere, na Hichapụ arụmọrụ n'ime otu okwu. Dịka ọmụmaatụ, iji MERGE, ịnwere ike ijikọ tebụl abụọ site na ịtinye ndekọ na-efu na imelite ndị dị adị. Jikota n'ime customer_account ca iji recent_transactions t NA t.customer_id = ca.customer_id MGBE NDỤMỌDỤ MGBE emelite itule = itule + azụmahịa_uru ma ọ bụrụ na ọ na-adakọghị wee tinye (customer_id, itule) Uru (t.customer_id, t.);a.
- Algorithms maka nhazi data na ebe nchekwa na diski emeziwanye nke ọma. Dabere n'ụdị data, ule na-egosi mmụba nke nhazi ọsọ site na 25% ruo 400%.
- Arụla ọrụ mpio site na iji row_number(), rank(), dense_rank() na agụta() dị ngwa.
- Emejuputala ohere nke mmezu nke ajụjụ ọnụ na okwu "Họrọ DISTINCT".
- Usoro maka ijikọ tebụl mpụga Data Wrapper (postgres_fdw) na-akwado nkwado maka mmemme asynchronous na mgbakwunye na ikike agbakwunyere na mbụ iji hazie arịrịọ na sava mpụga.
- Agbakwunyere ike iji LZ4 na Zstandard (zstd) algọridim iji mpikota onu ndekọ azụmahịa WAL, nke, n'okpuru ụfọdụ ọrụ, nwere ike melite arụmọrụ n'otu oge ma chekwaa ohere diski. Iji belata oge mgbake mgbe ọdịda gasịrị, agbakwunyela nkwado maka iweghachite ibe ha na-egosi na ndekọ WAL.
- Agbakwunyela nkwado maka ịpịkọta faịlụ ndabere na sava mpụga na ọrụ pg_basebackup. ihe nkesa, site na iji usoro gzip, LZ4, ma ọ bụ zstd. Ike iji modulu nchekwa omenala dị ugbu a, na-ewepụ mkpa ọ dị ịgba ọsọ iwu shei.
- Agbakwunyela usoro ọrụ ọhụrụ maka eriri nhazi site na iji okwu oge niile: regexp_count(), regexp_instr(), regexp_like() na regexp_substr().
- Agbakwunyela ikike iji chịkọta ụdị dị iche iche ("multirange") na ọrụ range_agg().
- Ọnọdụ nchekwa_invoker agbakwunyere, nke na-enye gị ohere ịmepụta echiche na-agba ọsọ dị ka onye na-akpọ oku kama ịbụ onye okike.
- Maka ntugharị ezi uche dị na ya, emejuputala nkwado maka nzacha ahịrị na ịkọwapụta ndepụta nke kọlụm, na-enye ohere n'akụkụ onye na-ezipụ ka ịhọrọ obere data site na tebụl maka mmegharị. Na mgbakwunye, ụdị ọhụrụ ahụ na-eme ka njikwa esemokwu dị mfe, dịka ọmụmaatụ, ọ ga-ekwe omume ugbu a ịwụpụ azụmahịa na-emegiderịta onwe ya wee gbanyụọ ndenye aha na-akpaghị aka mgbe achọpụtara njehie. Ntugharị ezi uche dị na ya na-enye ohere iji mmemme nkeji abụọ (2PC).
- Agbakwunyela usoro ndekọ ọhụrụ - jsonlog, nke na-echekwa ozi n'ụdị ahaziri site na iji usoro JSON.
- A na-enye onye nchịkwa ikike inye ikike nkeonwe nye ndị ọrụ iji gbanwee ụfọdụ paramita. nhazi sava PostgreSQL.
- Utility psql agbakwunyela nkwado maka ịchọ ozi gbasara ntọala (pg_settings) site na iji iwu “\dconfig”.
- A na-ahụta iji ebe nchekwa nkekọrịta maka ịchịkọta ọnụ ọgụgụ gbasara ọrụ nkesa, nke na-eme ka o kwe omume iwepụ usoro dị iche iche nke ịnakọta ọnụ ọgụgụ na ịtọgharịa steeti ahụ na diski oge.
- Enyerela ikike iji mpaghara ICU ndabara "ICU Collation" na mbụ, naanị mpaghara libc ka enwere ike iji dị ka ebe ndabere.
- Atụpụtala pg_walinspect ndọtị arụnyere, nke na-enye gị ohere iji akwụkwọ WAL nyochaa ọdịnaya nke faịlụ site na iji ajụjụ SQL.
- Maka atụmatụ ọha, ndị ọrụ niile, ewezuga onye nwe nchekwa data, enweela ikike ịmebie iwu CREATE.
- E wepụrụ nkwado maka Python 2 na PL/Python Ụdị nkwado ndabere na mpaghara anaghị akwụ ụgwọ.
Mgbakwunye: Site na 19:00 ruo 20:00 (MSK) a ga-enwe webinar na-ekwurịta mgbanwe na nsụgharị ọhụrụ na Pavel Luzanov (Postgres Professional). Maka ndị na-enweghị ike ịbanye na mgbasa ozi ahụ, ndekọ nke akụkọ Pavel's June "PostgreSQL 15: MERGE na ndị ọzọ" na PGConf.Russia na-emeghe.
isi: opennet.ru
