Shulou(Shulou.com)06/01 Report--

The emergence of the epidemic in 2020 has had a serious impact on the operation of many enterprises. In the face of emergencies, Giant Sequoia made use of the long-term accumulation of remote R & D cooperation system, still persisted in technological innovation, promoted R & D work in accordance with the existing plans, and officially launched Giant Sequoia Database (SequoiaDB) v5.0 Beta version.

We are also here to tell you what exciting features and features will be included in SequoiaDB v5.0.

Official support for the ARM architecture

Since version 3.2, SequoiaDB has supported ARM chip servers and domestic operating systems in limited versions. Since SequoiaDB v5.0, we officially support ARM chip architectures such as Feiteng and Huawei, as well as a variety of domestic operating systems, including UOS, depth, winning Kirin, and so on.

New monitoring and operation and maintenance system

For a long time, the design of the monitoring and operation and maintenance system of SequoiaDB is based on the bottom diagnostic snapshot (snapshot) call, which returns the operation information of the whole or part of the cluster to the upper layer application through Restful API. Generally speaking, enterprise customers will develop interface modules according to their own needs, and access the monitoring of SequoiaDB to their own monitoring large screen system or Zabbix, ELK and other kinds of software.

However, we find that it is difficult for some users to use this mechanism, and more planning and design and customized secondary development are needed. Therefore, in SequoiaDB v5.0, we have introduced an enhanced out-of-the-box graphical monitoring system in SAC (Sequoia Administration Center), so that users can use perfect graphical operation and monitoring capabilities without any pre-configuration.

The new graphical monitoring module in SAC is based on the open source project Prometheus and Grafana. In addition to displaying standard performance monitoring indicators, we can also support capabilities such as slow query Analysis (SequoiaPerf), so that users can analyze the history and current slow queries of the system in a full graphical interface.

​

Enhanced MVCC capability

MVCC multi-version control is one of the standard functions of database in recent years, and it is also an important basic capability for distributed database to achieve multi-interval RR isolation level. SequoiaDB supports MVCC multi-version control for the first time since v2.6, and has been further enhanced in v5.0.

In SequoiaDB v5.0, clusters introduce the concepts of global logical time (ULT:Universal Logical Timestamp), local logical time (LLT:Local Logical Timestamp), local UTC time (LRT:Local Real Time), and so on. The global unified time is synchronized periodically through the STP (SequoiaDB Time Protocol) protocol between nodes to avoid the performance loss caused by accessing the global time for each query.

After the initial protocol exchange, there is no direct relationship between the logical time error between nodes and the local physical clock of the operating system, that is to say, any operating system time jump will not have any impact on the cluster logical time of the giant sequoia database.

For a long time, SequoiaDB has implemented multi-version control MVCC using the RBS (RollBack Segment) rollback segment mechanism. In v5.0 version, the record information in the rollback segment is correspondingly introduced into the ULT field, which further enhances the distributed consistency transaction capability.

Intelligent problem diagnosis and analysis ability

For a larger cluster, it is quite complicated to diagnose problems. For example, due to some slight differences in system time, for hundreds of millisecond query performance diagnostics with high concurrency, tracking the same query session between multiple systems is quite complex.

Therefore, in a large cluster, the database must have a strong self-problem diagnosis and repair ability in order to minimize the difficulty of DBA and operation and maintenance personnel in diagnosing problems.

In the version of SequoiaDB v5.0, the problem location and diagnosis mechanism between multiple servers is rearranged in an all-round way, so that users can "see" the root causes of problems and faults more intuitively, which greatly reduces the difficulty of tracking and diagnosing faults and problems for operation and maintenance personnel.

Automate the highly available deployment of database instances

In previous versions, although SequoiaDB implemented the default high availability and disaster recovery mechanism for underlying storage, for the high availability solution of database instances created by upper-level users, the DDL log replication tool needed to be manually configured and could not be used right out of the box.

In SequoiaDB v5.0, users can directly specify instance groups when creating database instances such as MySQL, PostgreSQL, MariaDB, etc., and the DDL information of the database table structure in an instance group will be automatically synchronized to avoid additional manual configuration.

In addition to the major technical features mentioned above, there will be more new features in SequoiaDB v5.0. At the same time, there will also be a new large-scale improvement in database performance and security. I hope you will continue to pay attention!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.