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

The following data are taken from April 2019. For more information, please refer to http://db-engines.com/en/ranking/. The data are for reference only.

DB-Engines rankings are based on five different factors:

Number of keyword searches for Google and Bing search engines

Number of searches on Google Trends

Number of job searches on the Indeed website

The number of profiles with keywords mentioned in LinkedIn

Number of related questions and followers on Stackoverflow

News flash

1. On December 5, 2018, Berkeley RISE Lab launched a multi-data stream real-time distributed analysis system Confluo. It can be used as a network monitoring and diagnostic framework, as a time series database and publish subscription message system: as a time series database, its performance is several times higher than other time series databases, and as a publish message subscription system, its throughput is 4 to 10 times higher than that of Kafka.

2. RedHat removed MongoDB from its RHEL 8.0 Beta in mid-early January 2019 because, in October 2018, MongoDB announced that its open source license had been switched from GNU AGPLv3 to Server Side Public License (SSPL), hoping to make revenue from the software-as-a-service or cloud computing business, but changing the license is driving major Linux distributions away from MongoDB.

On February 12, 2019, Red Hat announced that Red Hat Satellite's underlying database would abandon MongoDB and switch to PostgreSQL.

3. On January 16, 2019, Google, together with researchers from MIT and Brown University, jointly launched a new database system, SageDB, and wrote a paper detailing the design principle and performance of SageDB. In their paper, they put forward a point of view: the learning model can permeate all aspects of the database system. "the components based on machine learning can completely replace the core components of the database system, such as index structure and sorting algorithm." even query executors. "

4. On January 24, 2019, Confluent, the company behind the open source Apache Kafka project, announced on its official blog the financing of D-round, valued at about $125 million, with a total valuation of $2.5 billion.

5. Microsoft acquired Citus Data on January 24th, 2019, reaffirmed its commitment to open source and accelerated the performance and expansion of Azure PostgreSQL.

6. On February 19, 2019, Redis Labs announced a new round of financing of $60 million. To date, Redis Labs has raised a total of $146 million. The new round of financing will be used to accelerate the market development strategy and continue to invest in the Redis community and product development.

On February 21, 2019, Redis Labs again changed its open source license to compete with cloud vendors, but it will not affect Redis's core license.

7. On March 11, 2019, F5 and NGINX announced a final agreement under which F5 will acquire all outstanding shares of NGINX with a total value of approximately $670 million.

8. On March 11, 2019, Nvidia came close to a deal to buy Mellanox Technologies Ltd for more than $7 billion in cash. The latter offers a wide range of data center products, including Ethernet switches, chips, and InfiniBand, while Mellanox accounts for more than 70 per cent of the InfiniBand market.

9. On March 19, 2019, Istio released version 1.1, focusing on and improving performance and scalability; data plane and control plane efficiency; namespace isolation; and multi-cluster functionality and availability.

10. On March 21, 2019, Ali Yuntu database GDB was officially tested, which is used to handle highly connected data query and storage, supports attribute graph model, and is fully compatible with TinkerPop Gremlin query language.

In March 2019, SequoiaDB Giant Sequoia Database officially joined the Global Top Open Source Community Cloud Native Application Computing Foundation (Cloud Native Computing Foundation, hereinafter referred to as CNCF) and became a member of the CNCF Foundation, one of the earliest open source cloud native databases in China. At the same time, based on its contribution to the open source cause, Giant Sequoia Database was also invited by CNCF to join the Linux Foundation.

RDBMS family

MySQL releases official version 8.0.15

On February 1, 2019, MySQL released the official version 8.0.15.

In addition to fixing bug and improving the quality of MySQL, the new features that are worth paying attention to are:

InnoDB supports parallel clustered indexes and improves the performance of CHECK TABLE to check tables for errors. It allows us to read clustered indexes in parallel with multiple threads.

In the previous MySQL, when the master database went offline, the group replication failure was transferred to the new master database. If the new master database is still digesting transactions synchronized from the original master database in relay log before entering the database, it will begin to accept new transactions sent by the client, which will cause new transactions not to find these changes that have not been reflected in the new master database, and cannot find the rows to be updated or deleted, thus causing synchronization inconsistency problems to emerge. You can now set the group_replication_consistency variable to BEFORE_ON_PRIMARY_FAILOVER so that the new master library will wait until all transactions in relay log enter the database before starting to process new transactions to solve the above consistency problem.

Group replication supports IPv6 in addition to IPv4, as well as mixed use of IPv4 and IPv6 among team members to support the process of upgrading from IPv4 to IPv6.

The ALTER TABLE command can change the character set in place.

Static binary log encryption, which can now encrypt binary and relay logs when stored on disk. Encryption is optional and can be enabled through the system variable binlog_encryption. With this new feature, users should be able to instruct the database to encrypt data stored in binary log files on disk. So that the data files have more comprehensive protection.

New features in MariaDB version 10.3

Since MariaDB10.2.7 (included) and above, online physical hot backup using Percona XtraBackup tool is no longer supported.

MariaDB 10.1 introduces features unique to MariaDB, such as InnoDB page compression and data at rest encryption. These exclusive features are very popular among MariaDB users. "however, existing backup solutions from the MySQL ecosystem, such as Percona XtraBackup, do not support full backup capabilities for these features."

In order to meet the needs of users, MariaDB officially decided to develop a backup solution that fully supports the unique features of MariaDB. It rewrites the extension based on Percona XtraBackup version 2.3.8.

1. Mariabackup Hot backup tool

The mariabackup tool is used (included in the binary tar package bin directory):

1) backup

Shell > mariabackup-defaults-file=/etc/my.cnf-S / tmp/mysql3306.sock-backup-target-dir=/data/bak/-user=root-password=123456

2) restore incremental data during backup (this step is similar to innobackupex-apply-log)

Shell > mariabackup-- prepare-- target-dir=/data/bak/

Note: / data/bak/xtrabackup_binlog_pos_innodb file, record the POS point of the main library show master status.

3) restore

Steps

Close the mysqld process

Make sure the datadir directory is empty. You can rename the original directory and create a new one.

Restore

Shell > mariabackup-defaults-file=/etc/my.cnf-copy-back-target-dir=/data/bak/

Change directory properties

Shell > chown-R mysql:mysql / var/lib/mysql/

If you want to back up on the slave library and want to record the POS point of the show slave status of the slave library so that you can pick up another slave library later, you can do this:

Shell > mariabackup-defaults-file=/etc/my.cnf-S / tmp/mysql3306.sock-backup-slave-info-safe-slave-backup-target-dir=/data/bak/-user=root-password=123456

Note: / data/bak/xtrabackup_slave_info file, record the POS point of show slave status from the library.

PostgreSQL releases version 11.2

PostgreSQL released version 11.2 on February 14, 2019. Compared with turning off parallelism, the average performance of 36 parallel computing scenarios is improved by 20 times.

1. PostgreSQL plug-in dynamics

Distributed plug-in citus release 8.2.1

Citus is a sharding plug-in of PostgreSQL. At present, there are a large number of use cases of SUNING, Tiezhong and Exploration in China.

Geographic information plug-in postgis releases 2.5.2

Version 2.5.2 fixed 23 BUG and relied on Proj 6 for the first time.

Timing plug-in timescale releases 1.2.2

Timescale is a time series database plug-in of PostgreSQL, which has a very good application in the IoT industry.

Stream Computing plug-in pipelinedb releases 1.0.0.13

Pipelinedb can aggregate the high-speed written data in real time according to the defined aggregation rules (support probability calculation), and trigger events in real time according to the defined rules (support the customization of event handling functions). It can be used in IoT, monitoring, FEED real-time computing and other scenarios.

2. PostgreSQL derivative open source products

Professional map database agensgraph releases version 2.1.1

Gpdb release 6.0 beta version of mpp database suitable for OLAP and HTAP scenarios

NoSQL family

Redis releases version 5.0.4

Redis released version 5.0.4 on March 18, 2019. This version fixes many bug, especially redis stream applications, that can cause Redis to crash in extreme cases.

Some important bug fixes in this release are as follows:

Fix Hyperloglog code errors that could lead to crashes of bug

Fix copied bug, which may crash in the event of a handshake command error

Fix bug with memory leaks caused by misuse of API

Fixed the problem that the ordered set ZPOP* count is 0

Resolution of redis-check-aof overflow problem

Improve LFU domain management

In mid-late February 2019, RediSQL released version 0.8.0.

RediSQL is a Redis module that provides complete SQL functionality for Redis. RediSQL supports all Redis data types and functions, as well as SQL statements SELECT / INSERT / UPDATE / DELETE and so on, as well as some nested join and data warehouse features (such as foreign key constraints).

This release introduces two new commands:

REDISQL.QUERY.INTO [.NOW]

REDISQL.QUERY_STATEMENT.INTO [.NOW]

The new command behaves like REDISQL.QUERY and REDISQL.QUERY_STATEMENT, but they XADD the result to the Redis stream as the first parameter.

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.