In this release we're releasing auto-mapping mode, Google Analytics integration, multi-table support for Postgres RDS, S3 CSV header detection, and more!
- Auto-mapping mode - We're super excited to announce one of our most killer features yet - auto-mapping mode! Are you tired of getting errors every time new fields get added to your data, requiring you to log in to Alooma to map and restream? You'll be happy to hear we added a mode that will catch when new fields are added to your events, map them based on our advanced data detection heuristics, and restream them automatically!
- You can set this as the default behavior for all event types from our Mapper Configuration page by clicking the gear icon on the Mapper (next to the sort option).
- You can override the default behavior on a per-event-type basis by clicking the gear icon next to each specific event type.
- Keep in mind that this setting works best for event types with regular data types but that do not have a strict schema, meaning events that are generated from inputs such as Segment or Mixpanel.
- For event types that come from schema'd inputs like MySQL, stay tuned, this is coming soon!
- If you have any questions about using our new auto-mapping mode, feel free to reach out to us at email@example.com and we'll be happy to help!
- Better Data Type Detection - Speaking of our advanced data detection heuristics, they've gotten even better! In this release we've improved our detection of:
- Boolean, which will be set if your data has the contents of "true" or "false", with any capitalization.
- Timestamp, which was improved to include integers indicating seconds since epoch, as well as to include additional string formats.
- We're also better at telling whether your integer timestamp is in milliseconds or seconds, so you always get the right conversion into Redshift.
- We now differentiate between floats and integers.
- Varchar lengths, which will now be set based on the next power of 2 longer than the longest string we detect in your new data instead of defaulting to 256.
- Array to Varchar Conversions - On the topic of auto-mapping and data detection, we've improved our auto-mapping logic to automatically convert arrays to varchars, so they'll no longer appear as invalid. If you'd like to perform any other parsing logic on your arrays, you can still do so in the transform function of the Code Engine.
- Postgres RDS Multiple Tables Support - If you're interested in adding your Postgres RDS data source into Alooma, that is now significantly easier! Our input now allows for adding as many table/replication index column pairs as you'd like, and the index column can be either a timestamp or a numeric type.
- Note that we now also automatically detects your Postgres RDS primary key for each table, so there's no need to define that anymore.
- Google Analytics Beta - You asked for it, and you got it - Google Analytics integration! You can start pulling your Google Analytics report data on a daily basis, based on Google's Core Reporting API. Just sign in with your Google account, select your dimension and measures, and go!
- Our integration is still in Beta, which means you'll have to contact us at firstname.lastname@example.org if you plan to use it, so we can add your domain to the list of authorized connections (otherwise, you'll get an authentication error).
- NAT Integration - The security of your data is of the utmost importance to us - that's why we've changed the way we communicate with your Redshift by putting our system behind a NAT. That way all communication comes through these dedicated IPs, and your Alooma deployment is safely behind it. For info on how to configure your Redshift to work with our NAT, check out our Granting Redshift Access documentation.
- We've expanded our S3 input to allow specifying whether the files are a delimited file with a header. That way we'll pick up the header automatically, and even let you explicitly define your delimiter, quote, and escape characters.
- Our type conversion errors, emitted when your event's field data doesn't match its associated Redshift column type, were expanded to be clearer and more specific. Now you'll know exactly what happened and how to fix it.
- We upgraded our MySQL input UI to now allow table names with capital letters.
- We've added a metadata field that represents the schema name to the MySQL metadata. This now allows streaming in tables with the same name from different MySQL schemas.
- That confusing auto-complete of username and password we were using for all of our inputs, that based itself off your Alooma login, was removed.
- We optimized the way we collect sample data to greatly lessen the chances of seeing an "Event overflow error", which affected many of your ability to map new fields when you had lots of large or varied data.
- Fixed an issue where the
$wififield for iOS events was always returning as false.
- Fixed the experience of the mapping type field dropdown not always opening on first click.
- Where once we had errors if you sent us events that had an input label with disallowed characters, now we just replace those characters with an underscore.