It would seem that the topic has been exhausted and a bright future may come for highly loaded bases, despite the fact that the 1C platform does not yet support these features from the “box”. However, we would like to dwell in more detail on such a topic as splitting a database into separate files using database management software. In the article on partitioning, database management software were already used for partitions, but there was said about them in passing. With the 먹튀디비 option you can now come up with the best options.
Today we will examine in more detail their use, as well as the nuances that must be reckoned with when servicing the database, restructuring and other points.
Why do you even need to split the database into separate files? The most common cases:
There is a register of information, which stores binary data of files. It is necessary to take out storage of files on a separate disk / storage in order to free up space on fast disks.
There are old archive tables that are rarely used, but data cannot be deleted. Why not even transfer such tables to separate disks, which are intended for this. Moreover, such database management software can be made read-only.
- Speed up database backup, because archive database management software cannot be backed up every time. They don’t change!
- Improved performance by distributing database files on separate media.
We will not discuss the topic of accelerating backups and performance now, but you can read about this in the publication on partitioning. We focus on the description of settings for file groups and their maintenance. All the examples below will be made for SQL Server, but for PostgreSQL this will work with some modifications.
Standard approach
Any database, whether for 1C or any other application, supports splitting the database into several files (of course, if it supports the DBMS). In the context of SQL Server, this is accomplished using database management software.
By default, the database contains only one file group “PRIMARY” , which 1C uses for its own purposes. In addition to tables and indexes, this predefined group stores service information about the database, various headers, etc. Therefore, you cannot completely replace this group with another, it will always be present.
However, we can add our own database management software and use them for 1C tables, and the platform itself will not know about it.
For example, take the BSP demobase and create two new file groups in it.
But just adding database management software is not enough. You also need to add data files for which these database management software will be involved.
Nothing to do
According to the third option, quite a few examples are written on the network, so we will consider only the first two points. We will do all the examples on the “History of Address Objects” information register, which on the database side is represented by the _InfoRg4683 table with several indexes.
Forward through restructuring
- And so, first we’ll install the main filegroup one of those added above.
- Now you need to do the restructuring of the platform. There is no normal way to call it for our case, but we can:
- Add temporarily props to the table, and then start the restructuring.
- Completely restructure the base through Testing and Correction.
Both options look “not very”, and you are lucky if the need for restructuring appears just at that moment for other tasks. When using the Test and Fix tool, you will generally transfer all tables and indexes to the installed file group, therefore we will not consider this case at all. And here is an example with the addition of a temporary attribute – please.
