HP Integrity rx1620 1.30 GHz 3 MB Bundle Solution AB569A Leaflet
Product codes
AB569A
Data protection
Event data must be protected from being changed or destroyed by any source, including applications,
Event data must be protected from being changed or destroyed by any source, including applications,
users, and component failures. Destruction or modification of event data can result in non-compliance
with government regulation, increased security risk, or failures in system management. Although
with government regulation, increased security risk, or failures in system management. Although
relational data has similar data protection requirements, in many cases these requirements are not as
absolute, and protection failures do not have the same pervasive impact.
Table1: Contrasting relational data with event data
Area of concern
Event data
Relational data
Transactions
Limited transactional requirements Fundamental
requirement
Stored data is seldom updated so all stored data is
available to all users (subject to authorization
filtering).
available to all users (subject to authorization
filtering).
Must present an isolated virtual database
view to prevent visibility of non-committed
data.
view to prevent visibility of non-committed
data.
Isolation concurrency
Must be general and flexible to accommodate
future event types. Semantics of data are often
determined at search time.
future event types. Semantics of data are often
determined at search time.
Generally static and must be determined
before data can be stored and accessible
by applications.
before data can be stored and accessible
by applications.
Schema
Search criteria are precise and databases
are optimized to support a known set of
queries. Search requirements are static or
evolve slowly.
are optimized to support a known set of
queries. Search requirements are static or
evolve slowly.
Search criteria can be precise or pattern based and
search requirements evolve rapidly.
search requirements evolve rapidly.
Search
Time attribute is part of every event and is usually a
key criterion for search.
key criterion for search.
Time is one of many possible attributes
and may not be present.
and may not be present.
Time attribute
It has no consistent life cycle
requirements. Most data has an indefinite
life span.
requirements. Most data has an indefinite
life span.
All event data is eventually removed or archived
based on aging and retention rules.
based on aging and retention rules.
Life cycle
Data protection to prevent destruction and
modification is absolute and must be supported
regardless of use access or component failure. Data
protection failures may cause compliance failures.
modification is absolute and must be supported
regardless of use access or component failure. Data
protection failures may cause compliance failures.
Data protection is customized per
application and is based on database
security authorizations and application
business logic.
application and is based on database
security authorizations and application
business logic.
Data protection
Lack of availability is likely to impact the entire
system and may have legal ramifications.
system and may have legal ramifications.
Lack of availability may only impact a
segment of the system.
segment of the system.
High availability
Load rate must keep pace with the event data
creation rate. System components generating event
data do not wait for event data loading.
creation rate. System components generating event
data do not wait for event data loading.
Load rate is based on user response time.
Increased response time reduces the load
rate requirements.
Increased response time reduces the load
rate requirements.
Load rate
Storage barriers of RDBMS-managed event data
The mismatches between event data management requirements and RDBMS core technology strengths
can be roughly depicted as either:
• RDBMS over-head is not required to manage event data or
• Lack of technology to support the unique requirements of event data management.
• Lack of technology to support the unique requirements of event data management.
Industry experience and empirical evidence show that these differences are manifested in real failures
to meet compliance, security and system management requirements. These failures can result in failing
to meet audits, increased security risks, and increased IT infrastructure costs.
There are multi-faceted problems in using an RDBMS to manage event data. Companies using RDBMS
There are multi-faceted problems in using an RDBMS to manage event data. Companies using RDBMS
technology in this manner are likely to experience most, if not all, of these problems.
7