Deep Dive on ClickHouse® Sharding and Replication | Tutorial for ClickHouse®

Подписаться 2,9 тыс.

Просмотров 8 тыс.

50% 1

Have questions about ClickHouse®? Get a free consultation with Altinity.
No marketing BS. no sales talk during the call. Just answers to your questions. Book a free help session: hubs.la/Q02qr1RZ0.
_______________________________
Join us as we dig into ClickHouse sharding and replication, showing how they enable clusters that deliver fast queries over petabytes of data. We start with basic definitions of each, then move to practical issues. This includes the setup of shards and replicas, defining schema, choosing sharding keys, loading data, and writing distributed queries. We finish with tips on performance optimization.
========================
PRESENTATION SLIDES
========================
* altinity.com/w...
==================================
MANAGED CLOUD FOR CLICKHOUSE
==================================
* altinity.com/m...
====================
RESOURCES
====================
* Altinity Knowledge Base for ClickHouse: kb.altinity.com/
* Altinity Documentation for ClickHouse: docs.altinity.com
* Altinity Stable Builds for ClickHouse: altinity.com/a...
* Training for ClickHouse: altinity.com/c...
* Altinity vs other ClickHouse vendors: altinity.com/a...
====================
COMMUNITY
====================
🧑🏽‍💻 Join the Community: www.altinity.c...
🙋🏽 Get Support on Slack: www.altinity.c...
⭐️ Star on GitHub: github.com/Alt...
🌐 Connect on LI: / 10955938
🌐 Follow on 𝕏: / altinitydb
🌐 Say Hi on Reddit: / clickhouse
#ClickHouse #sharding #dbmstutorials #Altinity #opensourcedatabase #olap #dbaas #opensourcecommunity #opensource #realtimeanalytics #datawarehouse #developers #devops #datamanagement

Опубликовано:

30 сен 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 7

@Altinity 7 месяцев назад

Want to be the smartest ClickHouse developer in the room? Then sign up to our newsletter to get ClickHouse tips and tricks straight to your inbox. Sign up here: hubs.ly/Q02pp8vh0

@Altinity Месяц назад

See an updated webinar on this topic here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-cTvjfOtX2S0.html

@maheshc1417 Год назад

in distributed databases clickhouse given we have cluster with 2 shards distributed factor of 2 and replication factor of 2 how many nodes can be down while the data being fully available? please advise on this or provide your answer for it

@roberthodges1480 Год назад

As long as you have a replica for each shard your database is up. It's just a copy of data like MySQL or PostgreSQL replication. For better availability consider 3 replicas--then if 1 is down you still have plenty of capacity to serve queries *and* replicate to a replacement copy.

@101graffhead Год назад

love the blog about clickhouse stuff, this all info about sharding is going to help me a lot in the inevitable migration we are gonna do, no one knows much beside querying clickhouse here. you people's love toward sharing information related to clickhouse is kind of peculiar but welcome regardless, almost none of the channels/blogs have this much information on clickhouse, also what are you guys thoughts on clickhouse keeper, its a new(at least new to me) alternative to zookeeper which is not written in java.

@Altinity Год назад

Hi! First of all you completely made my day with this comment. (This is Robert the person who did the video.) Second, ClickHouse Keeper is getting better but it's still not quite a complete replacement for Zookeeper. You can read our assessment here: kb.altinity.com/altinity-kb-setup-and-maintenance/altinity-kb-zookeeper/clickhouse-keeper/ Basically if you have an existing system that runs Zookeeper, we recommend you stick with it. If you are developing a new system that's a good time to go with ClickHouse Keeper. ClickHouse Keeper is also very handy for development, because it can run inside ClickHouse. You can therefore have replicated tables when you are developing on a laptop. This means you don't start out with, say unreplicated MergeTree, and then have to convert it to ReplicatedMergeTree as you get closer to deployment.