09/10/2012
Companion for MS Analysis Server v4.1
AS Companion Version 4.1 has been released with the following new features and reports:
a. The “Dimension Health Check” report has been added under the “Tuning Opportunities” node. The user can check if a dimension is healthy. The report checks if the declared attribute relationships match the underlying data. The attribute keys are verified for uniqueness. Potential new relationships are also reported. If the data source is not accessible, the user is prompted for a connection string.
b. The “Partition Overlap Check” report has been added under the “Tuning Opportunities” node. The user needs to select a partitioning key. The report checks if any of the partitioned members belong to more than one partition.
c. The “Dimension Statistics” existing report has been enhanced with Dimension selection and "Show records with zero count" option. This makes the report easier to read and interpret.
d. The ”Attributes With Default Member” is a new reported added under the “Metadata” node. The report lists all Attributes, which have default member other than “All”. This summary makes it easier for the cube developer to understand the default context of the cube queries.
e. The “Estimated Vs. Actual Counts” report has been added under the “Metadata” node. The report compares the declared in the cube estimated member count against the underlying data. Running this report can be expensive since it inquires the database about distinct counts for each attribute.
f. The “Many-To-Many Optimization Check” is a new report, added under the “Tuning Opportunities” node. The report checks if a cube contains a Many-To-Many relationships and for each one, it attempts to reduce the number of records in the intermediate table. The “M2M Optimizer” node allows the user to apply the proposed changes.
g. The “Object Usage Statistics” is a new report, added under the “Capacity Planning” node. It displays usage statistics for the different objects inside the database.
h. The “Roles” report has been added under the “Metadata” node. The report lists all the roles for the selected cube. For each role, the following fields are returned: Members, Allowed Set, Denied Set, and Default Member.
i. The “Object Processed State” report has been added under the “Metadata” node. The report lists the current state of each object inside the selected cube. The estimated size and rows are also included. The user has the option to filter out and see only the “Unprocessed Objects”.
j. The “Attach” event has been added to the list of events that can trigger the Cache Warmer. The “Attach/Detach” events are now reported under the "Backup and Restores" report
k. The “Data Type Discrepancy” report has been added under the “Metadata” node. The report list any objects, for which the underlying data source column has a different data type that the one used for the given attribute.
l. The “Non-Default Properties” report has been added under the “Metadata” node. It lists properties, which has been changed from their default value. This allows the cube developer to detect any possible issues and visualize the current state of the cube.
m. The “Dimension Usage” report has been added under the “Metadata” node. It lists all the relationships between the dimensions and the measure groups.
n. The “Dimension Optimization State” report has been added under the “Metadata” node. It lists the “Optimization” settings for each Attribute inside each dimension.
o. The “Aggregation Performance Check” report has been added under the “Tuning Opportunities” node. It allows the user to test the effectiveness of each Attribute Design and the associated partitions.
p. The "Object Memory Usage" report has been enhanced to present the statistics in an hierarchical form. The user can see on a high level what objects are taking the most memory and then drill down to the details. "Percentage of Total" memory usage column has been added as well. The report can be filtered to "Show Only Non-System Objects" or to show all objects that are taking memory. For performance reasons, only objects taking more than 0.99Mbs are included in the report. All other objects are grouped together in the entry: "Other".
q. The "Object Usage Statistics" report has been enhanced to present the statistics in an hierarchical form. The user can see on a high level what objects are used the most and then drill down to the details. "Percentage of Total" memory usage column has been added as well. For performance reasons, only objects with CPU and Reads greater than 0 are included in the report. All other objects are grouped together in the entry: "Other".
r. The "Processing Statistics" report has been enhanced to present the statistics in an hierarchical form. The user can see on a high level what objects are taking the most time and cpu to process. "Percentages" columns have been added as well.
