Shulou(Shulou.com)05/31 Report--

This article mainly shows you "what are the characteristics of HBase", the content is easy to understand, clear, hope to help you solve your doubts, the following let the editor lead you to study and learn "what are the characteristics of HBase" this article?

HBase is a database product of NoSQL, which is used to solve the problem of millisecond simple query of massive data.

Characteristics of HBase:

Table (table)-used to store data, data is also stored in the table.

Line (row)-- the term nesting.

Row keys (row key), similar to primary keys, are also used to distinguish each row.

Column cluster (column family)-- A collection of columns.

Column (column)-the equivalent of a field in RDBMS.

Cell (cell)-the cell used to store data.

Multiple versions of data-data with a timestamp.

The data in hbase is stored in a byte array, so the comparison is byte by byte.

Records in hbase are stored in rowkey order.

The column clusters in hbase are divided into many region according to their size, and different region can be stored in different region servers. that

Queries with large amounts of data can be converted into parallel queries.

Pre-partitioning operation means that when creating a table, it is divided into a specified number of region in advance, and the data is evenly distributed when inserting data.

In the same region.

In hbase, there is a table called meta. It stores the mapping between regionserver and rowkey of the user table.

In hbase, there is a table called-ROOT-, that stores .meta. The mapping relationship between rowkey and regionserver.

Represents a row of records in a table in hbase. See rowkey to determine a row of records.

There are many columns in the row, which are stored in many column clusters.

Each row and cluster is a separate storage file.

Some questions about Hbase:

Q: why does a row of records contain multiple timestamp values?

A: because this can reflect the idea of aggregation, it actually reflects the one-to-many problem in RDBMS.

Q: why is it that one row of records is stored in separate physical files according to column clusters, rather than storing them all together?

Answer: because when querying, not all fields are required for a row of records, only some fields are needed. If a row records the values of all fields together

In other words, it is obviously wasteful to read all the fields at once. If stored separately, only a small number of fields are needed for a query, so the query as a whole

The throughput will be improved.

Q: why is key regionserver,value rowkey and not the other way around when indexing?

A: because the number of rowkey is too large, if the index is built as above, the mapping data stream is too large.

Q: designing rowkey is the focus of hbase table design?

Answer: (1) because the hbase query is queried according to rowkey by default, the query conditions most needed by the business should be designed into the composition of rowkey.

(2) because the hbase table is partitioned, it is easy to produce hot spots when inserting data. In order to avoid hot spots, inserting data should be done as much as possible.

Random, requires the design of the main rowkey, preferably the start nodes of the rowkey are random or nearly random.

These are all the contents of the article "what are the characteristics of HBase". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

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.