Browse Prior Art Database

Method for Storage Optimized Historical Database

IP.com Disclosure Number: IPCOM000203478D
Original Publication Date: 2011-Jan-26
Included in the Prior Art Database: 2011-Jan-26
Document File: 5 page(s) / 695K

Publishing Venue

Siemens

Abstract

Software for managing and controlling Power Transmission and Power Distribution Networks usually provides capabilities for storing and accessing historical data in large archives. Some examples for data to be archived in the area of Power Transmission and Distribution are: • Measurement values of analog measurements, digital measurements and others • Accumulated values from accumulator measurements • Disturbance data which occur during network outages • Values manually updated by operators • Realtime Accumulator Processing values (RAP) • and others Raw values coming from field devices and stored in a database as “historical values” consume lots of disk space and are not needed by users. A user is usually interested in “aggregated values”, for example five minutes average values for all analog measurements or one-hour integral values of accumulator values. Therefore, database internal procedures usually aggregate the sporadic and non-cyclic raw values to a defined grid, e.g. five minutes or one hour. After having been successfully aggregated, the raw values are deleted from the database. Table 1 shows a simple example how such an aggregated data table can look like.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 38% of the total text.

Page 01 of 5

(This page contains 01 pictures or other non-text object)

(This page contains 00 pictures or other non-text object)

Method for Storage Optimized Historical Database

Idea: Bernd Steiner, DE-Nuremberg

Software for managing and controlling Power Transmission and Power Distribution Networks usually

provides capabilities for storing and accessing historical data in large archives.

Some examples for data to be archived in the area of Power Transmission and Distribution are:
• Measurement values of analog measurements, digital measurements and others
• Accumulated values from accumulator measurements
• Disturbance data which occur during network outages
• Values manually updated by operators
• Realtime Accumulator Processing values (RAP)
• and others
Raw values coming from field devices and stored in a database as "historical values" consume lots of

disk space and are not needed by users. A user is usually interested in "aggregated values", for

example five minutes average values for all analog measurements or one-hour integral values of

accumulator values. Therefore, database internal procedures usually aggregate the sporadic and non- cyclic raw values to a defined grid, e.g. five minutes or one hour. After having been successfully

aggregated, the raw values are deleted from the database. Table 1 shows a simple example how such

an aggregated data table can look like.

A database containing historical values usually contains a huge set of such archive tables, each for

different sets of aggregation cycles (e.g. 5 minutes, 1 hour, 1 day, etc.), different measurements types

(analogs values, accumulator values, disturbance data, etc.), different aggregations (average, integral,

minimum, etc.), and combinations of all of them (e.g. 5 minutes aggregated analog minimum, 5

minutes aggregated accumulator minimum, 5 minutes aggregated analog maximum, etc.).

Historical archived values, which exceed a certain age, are stored in such databases, even if they are

usually accessed rarely. However, it is obvious that storing all these data, for different aggregation

cycles, measurement types, aggregations and their combinations consumes lots of disk space. In order to avoid consuming lots of disk space, the following approaches were considered:
• Shifting "old" data to long term archive storages, such as tapes. However, if the data need to

be accessed, it requires lots of effort to access these data.
• Deleting of values that exceed a certain age. However, if customers insist on storing these

data, they must not be deleted.
• Storing data in low aggregation cycles, e.g. 1 day or 1 month. However, if customers might

need to generate one-hour reports, they will fail, as the high-cyclic archived values are no

longer available.

A new method for optimized storage of historical data in a database is proposed. It is illustrated in

Figure 2 and works as follows:
For each aggregation database table, an analogous so-called "Compression Table", an "U...