This document introduces new features included in Interana release 2.23 and lists newly resolved issues.
See Release 2.23.1 for information about the latest maintenance release.
Interana release 2.23 includes the following new features:
- .deb installer
- Cluster resize
- Global filters on dashboards
- Deduplicating query cache
- Data stats in column tooltips
- Enhanced arithmetic operations with derived columns
- Performance improvements
- System timezone consistency
- Unicode support in filters
- Streaming ingest API
- New integrations with Segment and mParticle
We've completely shifted our install method to a .deb package installer. This allows us to more easily distribute and install Interana software to supported Ubuntu environments.
Interana release 2.23 provides the ability to resize clusters for both data and string nodes.
Global filters on dashboards
You can now apply Global Filters to Dashboards. You can use them in the same way you use normal filters with Dashboards.
Deduplicating query cache
We re-architected our query cache with major performance improvements to dashboards. Unique queries will be deduplicated at the cache level, and refreshed based on intelligent prioritization. In addition, the following changes were made to dashboards:
- All charts now include the last updated time, which is shown relative to the current time (for example, "updated 3 minutes ago"). Interana will cache charts for up to 7 days; after that, Interana will automatically run a new query to refresh the charts. For email reports, Interana will cache data for 1 hour before running a new query. See Dashboards and query caching for more information.
- Use the Go button at the top of the Dashboard page to refresh all charts on the page.
Data stats in column tooltips
We've added data stats in column tooltips that provide context on your data, so you'll have more information to be able to select the right fields and named expressions to build your queries. You can see these in the Query Builder, and they include the following information:
|Occurrence||The percentage of events that have a non-null value for the column.
||The number of unique values in the column.
|Sample values||A list of sample values for the column.|
Enhanced arithmetic operations with derived columns
We added the
get function for derived columns. Use these when the data you're referencing contains null values, and you want a safe way to compute arithmetic operations on multiple fields.
This will work with both long and double values. The syntax for these is
long get(string field_name, long on_null); and
double get(string field_name, double on_null);.
See the Derived column syntax reference for more information.
We've made multiple enhancements to improve performance. This includes performance improvements for Dashboards and for string filters ("text contains," "starts with," "ends with").
System timezone consistency
In 2.23, we improved the way we store and reference timezones to ensure consistency across queries, named expressions, and dashboards. Interana references a system-wide timezone setting so all users have the same view of data, regardless of their browser timezone settings.
Unicode support in filters
Interana now supports unicode (UTF-8) text in filters.
|HIG-9498||Cannot share charts with email addresses that contain a plus (+) character.|
|HIG-9765||Incorrect embed code generated for charts when using Microsoft Edge.|
|HIG-9505||Cannot share a query for Distribution view that uses unpublished metrics.|
|HIG-9809||Must refresh to apply a column name change made from data tooltips in the Explorer.|
|HIG-10159||In the Explorer, scrolling doesn't work when the mouse pointer is on a table view.|
|HIG-10168||A dashboard that points to a non-existent table results in a confusing error message.|
|HIG-10170||Copying an invalid cohort is allowed, which then causes the UI to hang.|
|HIG-10180||Filter this out selects the contents of the entire table row instead of the column that was selected.|
|HIG-10194||Queries run while rebalancing return inaccurate results.|
|HIG-10268||If you re-use the name of a deleted table, the old derived columns appear in the new table as deleted.|
|HIG-10285||Admins are unable to edit the last column in datasets.|
|HIG-10289||Unable to make decimal columns aggregable.|
|HIG-10383||An edited metric does not update with new information until you perform a hard refresh.|
|HIG-10462||Newlines in advanced filter text break permalinks.|
|HIG-10485||Text no longer wraps for small dashboard charts, and cuts off the very bottom of the text for the first line.|
|HIG-10625||Import and query functions may temporarily be blocked during a string node rebalance. Once the rebalance is complete, performance returns to normal.|
|HIG-11026||In Explorer filters, if you exact match the name of a named expression, the named expression is added to the filter.|
Time syntax issues
The improvements to timezone consistency caused issues with some of our previously supported time syntax. These issues can potentially break existing named expressions and dashboards. We are currently investigating the scope of these issues and will be providing fixes for them in upcoming releases.
This maintenance release fixes the following issues:
|HIG-9858||Unable to parse datetimes.|
|HIG-10581||Queries returned "Query failed: map::at" error.|
|HIG-10732||Queries returned "no more data to read" errors after upgrading to 2.23.|
|HIG-10733||Global Filters are not saved when pinning charts to dashboards.|
|HIG-10810||Dashboard charts containing non-ASCII characters fail to load.|
|HIG-10811||The Text Contains expression does not support escape characters.|
|HIG-10822||The force_saml configuration parameter did not accept "enabled."|
|HIG-10823||String server is able to delete data in use.|
|HIG-10824||"No data to read" errors returned when large cache files created out of memory errors.|
|HIG-10853||Time syntax not parsed correctly for queries ending on the first day of the month.|
|HIG-10867||url_parse step does not recognize all valid URI schemes|