Release date: 26 October 2020
This page provides more information about the newly released functionality and improvements in UiPath Process Mining 2020.10.
UiPath Process Mining 2020.10 is focused entirely on improving the speed of the end-user experience in UiPath Process Mining. The whole Monitoring and Analyzing experience is much faster and more smooth as compared to earlier versions, as a result of:
- New Charts that load 5-12x faster than in 20.4 (bar charts, history charts, distribution charts).
- Improved caching that creates a faster and smoother end-user experience.
- Selector and metrics changes in your charts which are much faster than before.
Our benchmarks show that:
- With 50M records in AppOne, most dashboards load in under 5 seconds.
- With 100M records in AppOne, most dashboards load in under 10 seconds.
The loading speed of the dashboards is very dependent on the process graph. The variability of the process has a big impact on the loading speed.
Internet Explorer 11 will no longer be supported for UiPath Process Mining versions later than 2020.6. As an alternative, you can use Google Chrome, Firefox, or Microsoft Edge.
If you try to activate the 2020.10 build on Internet Explorer 11 you will not be able to use UiPath Process Mining. In this case, you must switch to another (supported) browser to continue using UiPath Process Mining.
Cache files use a lot of space on disk. To save disk space, it is strongly recommended to set the Compress contents to save disk space option for the UiPath Process Mining server installation folder. This will reduce disk space needed by about a factor 4.
See also Prerequisites for Installation
The new charts have a completely new User Interface that makes it easier to create them. The new charts also load up to 10x faster than their Legacy chart counterparts. Moreover, these charts can be cached in the disk cache, which makes sure that the initial loading times of these charts is significantly reduced as well.
We have new charts are available for:
- Bar charts;
- History charts
- Distribution charts.
Charts created with previous versions of UiPath Process Mining are now available as Legacy charts. Legacy charts will be deprecated in future versions of UiPath Process Mining. For Bar charts, History charts and Distribution charts, you are strongly advised to use the new chart types to get the associated performance benefit. Moreover, you should consider converting current legacy charts to new charts. For the chart types that do not have a new-style variant yet, you can still use their legacy versions, for example with Gantt charts, or detail lists or pie charts.
It is possible to convert legacy charts to new charts using the Convert to new chart option from the chart context menu. This converts the settings of the chart to the new chart.
It is possible to define metrics that determine the context of the main metric. In bar charts the context metrics are used to add columns to the chart. For all new charts the context metrics are visible in the hint when a user hovers over the chart.
For new charts it is now possible to define a drill-down attribute for a main attribute. Both the main attribute and the drill-down attribute must be present in an attribute selector on the chart. When a user clicks on the drill-down icon in the chart, the drill-down attribute is automatically selected in the attribute selector.
To drill-up the user can use the
Back button in the web browser.
Table items that have dynamic behavior do no longer have any caching. This includes at least usages of any of the following functions:
It is recommended to:
Today()for date/time. Alternatively, for
RealToday()only, it is also possible to add an expression attribute TableItem_RealToday with expression
RealToday(). You can use the attribute everywhere instead of the function. Note that this requires caches to be generated daily.
- Use a formula to generate predictable random sequences based on e.g. row number.
DateTimeActiveDatasource(). Note that this slightly changes behavior.
Also any table items with fatal errors do no longer have any caching. For example, circular references, broken references, syntax errors, etc.
In-memory caching is improved which makes it faster to go back to the previous filter state.
In AppOne the
tr function was used directly in all process graphs, which inhibits caching of these graphs. This is fixed in AppOne. It is recommended to also upgrade AppOne to version 2020.10. Otherwise you should remove the
tr calls from your process graphs and create a global expression attribute for the
tr call and referencing that instead in the process graphs.
The Flexible Heuristic Miner option is removed to further speed up the performance of the process graph. In previous versions of UiPath Process Mining this option was available on the Visualization tab of the Advanced Process Graph Settings.
The configuration file
.env should be renamed to the
tenant for which the configuration file contains the information. For example, if the tenant is
Default, the configuration file should be renamed to
Moreover, the information that must be specified in the configuration file has changed. For an Orchestrator on-premise installation the following information must be specified in the configuration file:
- instance type;
For an Orchestrator cloud installation the following information must be specified in the configuration file:
- instance type;
- user key.
It is now possible to change the tenant and the folder ID for which Actions and Queue items are created in the Basic Connector.
The tenant and folder ID must now be set in the
Globals table in the expressions
Folder ID. In addition, it is also possible to set a different folder ID per Action and Queue item that is defined in the Basic Connector.
Since a different folder ID can be set for each Action and Queue item,
folderID is now a mandatory field in the definition of Actions and Queue items.
reference is a mandatory field in the definition of a Queue item. The optional fields for Queue items now include the
postpone date. If
priority is not included in the Queue item definition, it is set to high by default.
reference can now be a unique reference to the Queue item. This makes it possible to create Queue items for queues with the option Unique Reference set in UiPath Orchestrator. When a Queue item is already present in a queue, it will not be created again. Duplicate queue items are allowed for queues without the option Unique Reference set.
The dashboards in AppOne are converted to the new charts for Bar charts, History charts and Distribution charts.
You might experience some small differences in headers or context metrics for example.
The dashboards in the Basic Connector are converted to the new charts.
In AppOne, on the Timing - Timing - Cases, Timing - Timing - Activities and Timing - Timing - Analysis dashboards, the metrics are combined into one metric selector, where there used to be two metric selectors.
On the Conformance - Tags dashboard in AppOne, the metric ‘Percentage of cases’ on the right dashboard item In which cases do tags occur has changed. When extra filtering is applied, such as filtering on a specific tag, the filtering is not applied to the computation of the percentage. This means that the percentage is calculated based on the total number of cases per row.
See illustration below.
For consistency reasons the ordering of tables in the Basic Connector and AppOne is improved. Also the construction of Cases filter and Unique record per ID in the BasicConnector - SingleFile is improved.
The Case department table item is added in the Basic Connector and in AppOne.
In AppOne it is now possible to drill down from Case department to Case owner.
The drilldown from Case type to Case owner is no longer available in AppOne.
Improved the construction of
Cases filter and the
Unique record per ID expressions in the BasicConnector - SingleFile. The
Cases filter and
Unique record per ID expressions are now combined which enables configurators to add filtering in the
Cases filter, which influences which records will end up in the
Cases_preprocessing table. The
Cases_preprocessing table no longer uses the
Cases filter directly."
Event ID (internal) expression in the Basic Connector which is used if the
Event ID attribute is not available in the input data. This makes it easier to define Due dates, when you do not have an
Event ID attribute available in the data.
The Process - Variants dashboard is no longer available in AppOne. As an alternative, you are suggested to use Process-Process-Cases dashboard and select Variant from the selector and/or use the Process - Compare dashboard and select Variant as the Compare attribute. The Variants dashboard is still available on the background and can be enabled by a developer if needed.
The following data-server functions are now deprecated, which means that they will be no longer available in UiPath Process Mining 2021.4:
If you use these functions you need to replace them by an alternative function. With UiPath Process Mining 2020.10 you can temporarily keep using these functions. In this case, you must prefix
-deprecated to the function name (e.g.
- In some cases, showing the filter results is much slower than expected;
- Changing the table is not enabled for new charts;
- The default chart metric has no name, which leads to floating sort icons;
- Selections do not work when a second attribute is added to an attribute selector;
- Column sort-state of new charts is missing:
- with Back and Forward in the browser history;
- when reapplying a Favorite.
- Selection states are inconsistent when having an unapproved selection:
- with Undo (CTRL-Z) and Redo (CTRL-Y);
- with Back and Forward;
- when reapplying a Favorite.
Updated 9 months ago