зеркало из
https://github.com/iharh/notes.git
synced 2025-10-30 21:26:09 +02:00
97 строки
5.3 KiB
Plaintext
97 строки
5.3 KiB
Plaintext
https://presentations.clickhouse.com/
|
|
https://presentations.clickhouse.com/2017-shad/index.html
|
|
|
|
2025
|
|
Otus - Dictionaries, Window and Table Functions 0:00 of 1:14:11
|
|
https://www.youtube.com/watch?v=iWrogMtdfjM
|
|
Otus - ClickHouse place in SystemDesign 0:00 of 1:53:23
|
|
https://www.youtube.com/watch?v=ksoa7EmhwoA
|
|
SqlNinja - Artemiy - What Is the Truth from What is Talking about ClickHouse of 1:05:58
|
|
https://www.youtube.com/watch?v=7ziBC_lb6TY
|
|
! CH is motly like LSM-tree (SS-Table), write-optimized (MergeTree engine)
|
|
! CH has no mem-table (data is going directly to disk)
|
|
! 5:00 internals (data storage)
|
|
! 7:00 We can't change anything in written chunks (append-only)
|
|
! => There are no locks in CH
|
|
! 8:00 Over time, this leads to huge not-needed data
|
|
! If we need to delete some data, then write a new chunk with thomb-stone, marking that some key is deleted
|
|
! 9:00 Over time we merge chunks into a bigger ones
|
|
! 10:00 OPTIMIZE TABLE <table-name>;
|
|
! 11:00 SELECT active, name, rows FROM system.parts WHERE database='test' AND table='table';
|
|
! 16:00 mutations (ALTER TABLE <table-name> UPDATE/DELETE WHERE ...)
|
|
! 18:00 mutations are faster than optimizations, but slow in general, non-consistent if dropped in the mean-time
|
|
! 21:00 async_insert setting is available at CH (around 1-sec buffer)
|
|
! 23:00 indexes consist of granulas (can represent multiple records each)
|
|
! 25:00 SELECT part_name, mark_number, rows_in_granule, idx FROM mergeTreeIndex('test', 'table') LIMIT 4;
|
|
! 32:00 SET send_logs_level='trace' SELECT ...
|
|
! 35:00 SELECT ... SETTINGS join_algorithm='full_sorting_merge';
|
|
! 38:00 CH and JOINs details
|
|
! 39:00 ENGINE=Dictionary !!!
|
|
! 40:00 Plus/Minus of CH,
|
|
! Super Cool SQL Dialect, Wide range of useful functions
|
|
Otus - Integration with Kafka, PG, S3 and Superset ru of 1:27:53
|
|
https://www.youtube.com/watch?v=GOT0kiOy-_M
|
|
! tg chat ???
|
|
! 1:07:00 !!! ... engine kafka settings ... ;
|
|
2024
|
|
Otus - ClickHouse Arch ru of 1:28:21
|
|
https://www.youtube.com/watch?v=xpn41LXVPu4
|
|
! 1:25:00 - list of resources
|
|
Otus - Kafka and ClickHouse of 1:41:20
|
|
https://www.youtube.com/watch?v=Jp1SEm4UU48
|
|
! 49:00 - ClickHouse Kafka Connect Sink
|
|
! 52:00 - REST Api for Kafka Connectors
|
|
! 1:00:00 - Debezium Docker image for pg
|
|
2023
|
|
Otus - ClickHouse Effective HighLoad Usage ru of 1:32:02
|
|
https://www.youtube.com/watch?v=d_LSJvBk_NU
|
|
ClickHouse - Kita - Modern SQL in 2023 20:00 of 33:05
|
|
https://www.youtube.com/watch?v=zhrOYQpgvkk
|
|
! 18:00 "explain query tree" system query
|
|
! ... run_passes=0/1 select id, value from test_table;
|
|
ListenIT - What ClickHouse is and Columnt DBs ru of 14:46
|
|
https://www.youtube.com/watch?v=2b4heaO3U8k
|
|
PostHog - In-depth: ClickHouse vs PostgreSQL of 7:32
|
|
https://www.youtube.com/watch?v=iLXXoDaFoxs
|
|
YandexCloud - ClickHouse Naughty Advices of 32:03
|
|
https://www.youtube.com/watch?v=UR0rmA_-KfA
|
|
! 17:00 !!! Do not insert by each single raw - every insert - is a separate part
|
|
2022
|
|
Altinity - High Performance, High Reliability Data Loading on ClickHouse® | Webinar for ClickHouse® of 1:00:52
|
|
https://www.youtube.com/watch?v=rexGRyl8yuA
|
|
Altinity Quickstart for ClickHouse® — Creating Your First Application | Tutorial for ClickHouse® of 1:07:02
|
|
https://www.youtube.com/watch?v=rawoPXXGiYY
|
|
! 19:02 show tables from system;
|
|
Altinity - Altinity Quickstart for ClickHouse® | Build Your First App | Training for ClickHouse® of 59:50
|
|
https://www.youtube.com/watch?v=phTu24qCIw0
|
|
! 48:00 storage internals - files for blocks, idx, ...
|
|
SmartData - Shutak - Insert into ClickHouse and not to die of 1:01:45
|
|
https://www.youtube.com/watch?v=XKBYYP5k_Uo
|
|
! 10:00-15:00 deduplication cfg
|
|
! 18:00 - insert_deduplicate (shocking default behavior)
|
|
! 40:00 - Kafka Engine and shards unballancing fight
|
|
!!! https://github.com/odnoklassniki/spark-to-clickhouse-sink
|
|
2021
|
|
Percona - Robert Hodges and Alexander Zaitsev - ClickHouse Developer Tutorial, Part 2 - Lab Exercises of 26:55
|
|
https://www.youtube.com/watch?v=j1vU43gbcd8
|
|
Percona - Robert Hodges and Alexander Zaitsev - ClickHouse Developer Tutorial, Part 1 - Intro to ClickHouse of 1:05:39
|
|
https://www.youtube.com/watch?v=WVk5_tQ-wx8
|
|
2020
|
|
HighLoad - Krasheninnikov - ClickHouse Data Processing Patterns ru of 51:47
|
|
https://www.youtube.com/watch?v=JSLhU_LJ5VE
|
|
https://highload.ru/spb/2019/abstracts/4862
|
|
2018
|
|
HighLoad - Zatelepin - What Need to Know about CH Arch ru of 51:02
|
|
https://www.youtube.com/watch?v=PLMSA_gDdyM
|
|
https://habr.com/ru/articles/509540/
|
|
https://temofeev.ru/info/articles/chto-nuzhno-znat-ob-arkhitekture-clickhouse-chtoby-ego-effektivno-ispolzovat-aleksey-zatelepin-2018g/
|
|
! 11:00 index writes stuff only for each 8192 rows
|
|
! 15:00 reordring via MergeTree engines (idea is like LSM-tree)
|
|
! 16:00 sorted data are written to disk in chunks (recommended no less than 1000 records)
|
|
! 17:00 background merge
|
|
! 18:00 if we have a lot of chunks, insert-perf will be really slow
|
|
! 20:00 sharding
|
|
YandexForDevelopers - Milovidov - SHAD - 005 - How ClickHouse Works ru 38:00 of 1:56:30
|
|
https://www.youtube.com/watch?v=vbhSrZxm66E
|
|
https://habr.com/ru/companies/yandex/articles/273305/
|