Pigsty 2.4: PG16 support, RDS monitoring and new extensions

Pigsty followed up with the new version Pigsty v2.4 after the release of PostgreSQL 16 , providing full support for the official version of PostgreSQL 16. In addition, in v2.4, additional support is provided for monitoring existing PG instances, especially RDS for PostgreSQL and PolarDB. Redis monitoring has been improved based on 7.x, providing automated Sentinel-based high-availability configuration.

Pigsty v2.4 is still in Beta status, you can use the following command to get started quickly. After the document repair is completed, it will be officially released.

bash -c "$(curl -fsSL https://get.pigsty.cc/beta)" 

Highlights

  • PostgreSQL 16 is officially released, and Pigsty provides support within 1 hour of release.
  • Can monitor cloud database, RDS for PostgreSQL, and PolarDB, providing a new PGRDS monitoring panel
  • Officially provide business support and consulting services. And released the first LTS version, providing subscriber customers with up to 5 years of support.
  • New extension: Apache AGE, providing graph database query capabilities on PostgreSQL
  • New extension plug-in: zhparser, Chinese word segmentation, used to support Chinese full-text search function
  • New extension plug-in: pg_roaringbitmap, efficiently implements the RoaringBitmap bitmap function
  • New extension plug-in: pg_embedding, another vector database plug-in based on HNSW index hnsw alternative to pgvector
  • New extension: pg_tle, a trusted language stored procedure management/publishing/packaging extension by AWS
  • New extension plug-in: pgsql-http, uses SQL interface to directly send HTTP request processing responses in the database.
  • Other new plug-ins: pg_auth_mon, pg_checksums, pg_failover_slots, pg_readonly, postgresql-unit pg_store_plans, pg_uuidv7, set_user
  • Redis improvements: Support Redis sentinel monitoring and configure automatic high availability of master-slave clusters.

API changes

  • A new parameter, REDIS.redis_sentinel_monitor, is used to specify the main library list for Sentinel cluster monitoring.

PG16 support

Pigsty may be the earliest distribution to provide PostgreSQL 16 support, starting from 16 beta1, so when PostgreSQL 16 was released one hour later, Pigsty completed support for the official version. You can already launch a high-availability cluster for PostgreSQL 16, although there are some important extensions that are not yet available in the official PGDG repository, such as Citus and TimescaleDB. But some other extensions are already available: including postgis34, pgvector, pg_squeeze, wal2json, pg_cron, and extension plug-ins maintained and packaged by Pigsty: zhparser, roamingbitmap, pg_embedding, pgsql-http, etc.

PostgreSQL 16 has some practical new features: logical decoding and logical replication from the database, new statistical views for I/O, fully connected parallel execution, better freezing performance, new function sets that comply with SQL/JSON standards, and Using regular expressions in HBA authentication and more.

However, it should be noted that the PGDG official repository has currently decided to abandon support for EL7 in PostgreSQL 16, so PG16 is only available in EL8, EL9 and their compatible operating system distributions.

Monitor RDS and PolarDB

Pigsty v2.4 provides support for RDS monitoring. In particular, monitoring support for PolarDB cloud database has been added. When you only have one remote PostgreSQL connection string, you can use this method to include it in Pigsty monitoring.

Pigsty v2.4 provides support for RDS monitoring. In particular, monitoring support for PolarDB cloud database has been added. When you only have one remote PostgreSQL connection string, you can use this method to include it in Pigsty monitoring. Pigsty provides two new Dashboards: PGRDS Cluster and PGINS Cluster, which are used to present complete indicators of RDS PG.

Commercial Support and LTS

Pigsty v2.4 is the first LTS version and will provide long-term support for enterprise subscribers for 3 years. At the same time, we will officially begin to provide external subscription and support services.

REDIS high availability

Pigsty v2.4 provides a new parameter redis_sentinel_monitor, which is used to automatically configure the high availability of classic master-slave Redis clusters. This parameter can only be defined on the Sentinel cluster. The main library in the definition will be automatically managed by the Sentinel cluster.

At the same time, Sentinel-related indicators and panels have been added to Redis monitoring, and adapted to the new features of Redis 7.x.

new extension

Pigsty v2.4 provides a series of new extensions, including important extensions that are not yet included in the official PGDG repository. For example, the graph database plug-in Apache AGE, the Chinese word segmentation full-text search plug-in zhparser, the HTTP plug-in pgsql-http, the trusted extension packaging plug-in pg_tle, the bitmap plug-in pg_roaringbitmap, and another alternative implementation of the vector database plug-in PGVector pg_embedding, etc. .

All plug-ins are compiled and packaged on EL7 - EL9 for PostgreSQL 12 to PostgreSQL 16. However, EL7 does not yet support pg_tle and pg_embedding due to compiler version issues. These RPM packages will be maintained by Pigsty and placed in Pigsty's own Yum source .

For example, you can use AGE to add graph database capabilities to PostgreSQL, create Graph, and use Cypher query language and SQL language to explore graph data to achieve the effects of Neo4j.

For another example, you can use the zhparser Chinese word segmentation plug-in to split Chinese text and queries into keywords, and use PostgreSQL's classic full-text search capabilities to achieve the effects of search engines and ElasticSearch.

What's more, you can also use the pgsql-http plug-in to use the SQL interface to send HTTP requests and process HTTP responses. This allows the database to be deeply integrated and interacted with external systems, opening up endless space for imagination:

You can also use roaringbitmap to efficiently perform counting statistics using very few resources:

Guess you like

Origin www.oschina.net/news/258249/pigsty-2-4-beta