The analytical views in Arcadia Data have four possible states:
UNUSABLE
, INVALID
, STALE
, and
FRESH
.
In addition to describing the Four States of Analytical Views, this topic explains how an analytical view may become stale, and what events trigger transitions between states.
State | Description |
---|---|
UNUSABLE |
Analytical view is created, but has never been refreshed. |
INVALID |
Base tables(s) were subject to DDL operation after most recent refresh. The analytical view is not consistent with base table definitions, and cannot be used. |
STALE |
Base table data was changed after most recent refresh. The analytical view may be used, but the query results may be incomplete/inaccurate. |
FRESH |
The data in the analytical view is current, or up-to-date with the base table(s). The analytical view may be used to safely route the query, and will return accurate results. |
In Arcadia Data, it is possible to control the staleness of analytical views at the session level.
The STALENESS_PERIOD
query option specifies, in minutes, how stale an
analytical view can become before ArcEngine stops routing queries to it.
The default value of STALENESS_PERIOD
is 1440 minutes, or 24 hours.
ArcEngine checks the difference between current time and the last successful refresh of the
analytical view to determine its age. If the age exceeds the STALENESS_PERIOD
value, ArcEngine marks the analytical view as stale
.
Various database events can change the state of an analytical view. Some operations
have no effect, while others trigger a state change to INVALID
.
Database operations that do not have an effect on the state of analytical views include the following:
INVALID
.INVALID
.ADD
operations on columns to the base table does not cause analytical
views to become INVALID
.Database changes that trigger an INVALID
state in analytical views
include the following:
DROP
or REPLACE
operation on the columns of the base
table that are used in the analytical view.