Treetable

Treetable massively enhances the usability of the BVQ GUI. It is a complete new face to the user, that justifies to give this release a new major number - 4.0.

It shows all attributes of all available objects ordered according the same path the Treemap uses. It is a replacement of the classic table view always visible below the Treemap with a synchronized selection.

Now the user has two perfect methods for configuration browsing: Treemap and Treetable.

Treemap stays to be the best tool for to get an overview about load or capacities of concurrent consumer type objects in a shared resource like a pool, storage class or node. It is perfect to show group relations across all objects in one picture. Selected objects show up in all of their groups, in some cases only as single pixel, but visible.

Treetable as an extension of Treemap adds capabilities with even more benefits:

• Perfectly suited for text based identification of objects
• Selectively shows all objects with all of their attributes grouped in categories
• Allows sorting and filtering
• Arranges the object hierarchy according to the same path as Treemap
• Keeps the selection of objects in sync with Treemap
• Can be exported via system clipboard or CSV file.

Here is a picture of the whole story:

How to work with it?

The Treetable will always open below the Treemap. A split bar between both allow you to resize both as needed or to enlarge one of both to fill the whole window.

Treetable arranges the object hierarchy according to the same path as Treemap. The last Object type configured in the path browser is shown with a selectable set of Attributes in the Treetable.

The selection of Objects is synchronized between Treemap and Treetable. All objects selected in Treemap will also be selected in the Treetable and vice versa. Collapsed folders in Treetable will automatically expand if one of their contained objects are selected in the Treemap. Treetable will try to show the most recent selected object in a visible area.

We invested months of work to reorganize and consolidate all the attributes and giving them appropriate short and long names and in many cases a helpful description. Hover your mouse pointer over a column header an wait for the tooltip. Long name and the optional description will be shown therein

With the Treetable menu you can select a singular attribute or a whole group of attributes to be shown as columns in the Treetable. To open this menu, you need to click on the button in the top rightmost corner of the Treetable or use the popup menu on one of the column headers.

The Treetable can be switched to a sort and filter mode. This can be initiated by selecting Sort and filter mode from the Treetable menu or with a double click on a column header.
In sort and filter mode you are able to set filters in the filterrow, or to sort the table rows according to the contents of columns. Simply click a single column header or hold the CTRL key pressed while clicking multiple to define a sort sequence.

The column containing the folder tree in normal mode is exchanged with a set of columns, one for each configured object type as configured in the current Path, representing the primary name of each object accordingly.
Such a table also fits perfectly to be exported to a CSV for further handling in a spreadsheet tool.

Treemap filtering

Ever analyzed very large clusters? Now you can concentrate the Treemap contents on one or more selected objects. Simply single click the object you want to isolate with the middle mouse button. You can also select Isolate from the object menu. If you selected more than one object in advance, Isolate will keep all selected objects in the upcoming focus. The filtering will also be effective in the Treetable.

A "*" character wil be displayed behind filtered objecttypes the path browser.

To get back to to an unfiltered Treemap, just single click the middle mouse button on filtered objects again or use the Clear filter entry in the path browsers objecttype menu.

Object centric navigation

You now can navigate directly and more easy from a selected object to related objects without the need to use the path browser for that. Simply select the objecttype you want to see grouped inside an object. For this selet the desired objecttype behind Isolate > Detail this in the object menu. The number behind each objecttype in the submenu is the count of related objects, that will be displayed inside. If there are no related objects for a specific type, it will not show up in the submenu.

Global analytic favorites

With the introduction of Analytic favorites in this release, favorites in general became more important than ever, so we also added some collaboration features to the favorite manager.

Collaboration features

New favorite management functions now allow to share them more easily between the users of inside a team and to define standard analysis processes with them.

The new main features of the favorite manager are:

• Favorites are exportable and importable
• Global favorites are stored in the BVQ DB and thus are available for all users. This allows collaboration between users and a continuous enhancement of the workflows supported and driven by BVQ.

A new favorite type called "Analytic Performance Favorite" adds major benefits:

Analytic Performance Favorites

The new Analytic Performance Favorites allow to store Performance view configuration settings more abstractly. With that property, they can be applied more flexibly to other objects or even other object types than initially defined. They allow to invoke a well defined performance view configuration on different objects in future situations.
This is most beneficial when multiple Analytic Performance Favorites are defined as part of a sequence of a comprehensive step-by-step analytic process.

In comparison, the classic BVQ performance favorites are more like an exact screenshot that can be recalled for exactly the same object set and timeframe showing a situation recorded in the past. Analytic Performance Favorites are decoupled from any time (duration) and object. Instead, it acts on the following currently active user selections:

• Selected objects in the related topology view
• Current timeframe, as most recently set in the performance view  (current adjustment of the performance time favorite)

They can be invoked in the following ways:

• From the favorite manager
• From the object menu embedded in a shortened stucture as defined in the favorite manager
  
  • Menu entry Favorites offers all Analytic Performance Favorites
  • Submenues of type specific performance view entries offer only Analytic Performance Favorites of the appropriate type.

Example of Analytic Performance Favorites in the favorite manager:

Example of Analytic Performance Favorites shown in the favorites:

Example of Analytic Performance Favorites shown behind type specific performance view menu entries:

Customize GUI startup window layout

You can save your current window layout to be reused automatically after each BVQ GUI start. Choose Main menu > BVQ > Save layout as default... entry to do this.

Consolidated DB installer

The two previously separate installers BVQ DB setup and BVQ DB schema migration are now consolidated to one tool. This will ease and accelerate upgrades and inital installations.

Pre-consolidated stats storage

Storing performance statistical data in the BVQ DB is now much more space efficient and data can be loaded much faster. Most counters are now aggregated across all cluster nodes and stored to DB after this pre-processing step. That will reduce the amount of data to be stored roughly by a factor equivalent to the number of nodes in the system. For example an 8 node cluster will only need 1/8 of the space for VDisk or MDisk data compared to previous versions of BVQ.

To overcome the need for a very long data conversion, the GUI and reporting modules have been enhanced to read both formats without any user intervention. Due to this capability, old data is kept until deleted by the BVQ schedulers space retention period (default: 100 days). Thus, the maximum space reduction is not expected to be seen at once. It can only be fully realized after the retention period.

Performance view enhancements

• Less system memory usage and much faster drawing in case of a massive high count of objects
• Manual rescale of every Y-axis
• Store chart configuration options configured in the chart menu in favorites or as default configuration reused at every start of a performance view
• Quick background color chooser menu integrated in the chart menu
  

Adaptive Treemap Sizing enhancements

Adaptive Treemap Sizing (ATS) is now used for all object attributes, that can be selected in the aspect selector of the topology view. That includes: Performance, cost and  price aspects.
An additional information in the text overlay of a Treemap object pad starting with the character ∑ shows the current object type used as aggregation source.

System deletion tool integrated in GUI

Discovered Systems (VMware, SVC, Storwize, ...) can now be deleted very easy. Just Open the main menu BVQ and select Delete discovered system... and follow the appearing wizard's instructions.