10.5.5 Data Backup
The only way to ensure data integrity is to have adequate backup strategies and methods. Types of data backup include
• Online mirrored backup. This type of backup requires an identical
amount of online storage that the databases are backed up to on a frequent
basis (often several times a day), preferably on a separate server located
at a different site than the original. Data libraries that have high use and
frequent updates should have this type of backup.
• Several disk drives on a single server that provide enough redundancy
so, if a hardware failure occurs, they can be reconfigured quickly to restore
the database. While they are technically not a backup system, they provide
a type of data security.
• Tape-archive hardware and software can be automated so the backup
occurs unattended. Various schemes for full and partial backups can be integrated to provide for the needs of the user. Adequate documentation about
the backup process is required. Some portion of the data backups should
be stored in a different place than the original database in case of fire or
other catastrophic event.
• High-capacity removable storage media (such as compact disk [CD] or
digital video disk [DVD]) provide the ability to store and back up data in
a form that is easily restored or accessed. They can often be accessed and
used more easily than tape backups but may require more operator intervention during their production than do tape backups.
10.5.6 Data Stewardship and Warehousing Groups
Historically, data were acquired, analyzed, interpreted, and reported by the
individual user. After publication, the data would be archived with little or
no use. With more modern technology for the storage of large quantities of
data, it is possible to warehouse the data in a database that can be easily
accessed by the user or shared with other users. Today, a great deal of data
is held by federal and state agencies, and much of it is available over the
Internet. The USEPA, U.S. Geological Survey (USGS), and other agencies
have hundreds of data sets, many of which are available for downloading
or accessible through direct database connectivity. In the future, users will
be able to access most data sets directly and may not even have a need to
keep local copies of widely available data except when more rapid access
is needed. A structured-query-language (SQL) database that is available
over the Internet is preferable in many ways to local copies of a database.
The primary data holder can maintain and update the files and access to
the data can occur when needed. Mirroring data from the primary holder
is useful because it can allow faster and easier access in many cases.
Mirrored data sets also act as a type of backup.
10. Effective Ecological Modeling: Data Issues
201
The only way to ensure data integrity is to have adequate backup strategies and methods. Types of data backup include
• Online mirrored backup. This type of backup requires an identical
amount of online storage that the databases are backed up to on a frequent
basis (often several times a day), preferably on a separate server located
at a different site than the original. Data libraries that have high use and
frequent updates should have this type of backup.
• Several disk drives on a single server that provide enough redundancy
so, if a hardware failure occurs, they can be reconfigured quickly to restore
the database. While they are technically not a backup system, they provide
a type of data security.
• Tape-archive hardware and software can be automated so the backup
occurs unattended. Various schemes for full and partial backups can be integrated to provide for the needs of the user. Adequate documentation about
the backup process is required. Some portion of the data backups should
be stored in a different place than the original database in case of fire or
other catastrophic event.
• High-capacity removable storage media (such as compact disk [CD] or
digital video disk [DVD]) provide the ability to store and back up data in
a form that is easily restored or accessed. They can often be accessed and
used more easily than tape backups but may require more operator intervention during their production than do tape backups.
10.5.6 Data Stewardship and Warehousing Groups
Historically, data were acquired, analyzed, interpreted, and reported by the
individual user. After publication, the data would be archived with little or
no use. With more modern technology for the storage of large quantities of
data, it is possible to warehouse the data in a database that can be easily
accessed by the user or shared with other users. Today, a great deal of data
is held by federal and state agencies, and much of it is available over the
Internet. The USEPA, U.S. Geological Survey (USGS), and other agencies
have hundreds of data sets, many of which are available for downloading
or accessible through direct database connectivity. In the future, users will
be able to access most data sets directly and may not even have a need to
keep local copies of widely available data except when more rapid access
is needed. A structured-query-language (SQL) database that is available
over the Internet is preferable in many ways to local copies of a database.
The primary data holder can maintain and update the files and access to
the data can occur when needed. Mirroring data from the primary holder
is useful because it can allow faster and easier access in many cases.
Mirrored data sets also act as a type of backup.
10. Effective Ecological Modeling: Data Issues
201