Snowplow RDB Loader 2.2.0 released

Announcements New releases
1

We’re happy to announce release 2.2.0 of Snoplow RDB Loader.

This release brings many new configuration options, making RDB Loader more flexible and several improvments in features we’ve introduced in recent 2.1.0 release.

  • Shredder has new runInterval section, allowing to use Shredder along with automatic S3 clean-up policy, as well as running parallel Shredders on different time windows
  • In RDB Loader, retries are now fully configurable
  • Loader can force restart the loading with pre-configured timeout if connection hangs
  • Health Check metrics introduced in 2.1.0 now get sent to StatsD
  • And other fixes and improvements in logging and retry queue

Upgrading

All mentioned settings are optional and 2.2.0 will be working with older configs without change.

5 Likes
2

I don’t want to sound ungrateful or impatient, but didn’t you say you were going to release the RDB Loader 3.0.0 with Snowflake and streaming this month? There are only two workdays left :grimacing:

3

Hi @medicinal-matt it’s great that you’re excited about 3.0.0 release – I am too!

We wanted to get this 2.2.0 release out first to round out the 2.x series with a robust and dependable loader, that will satisfy Redshift users for a good while longer. We have been working on the 3.0.0 release in parallel, and it is pretty much code complete, including loading into Snowflake.

Next week, I anticipate we will put out a “release candidate”, which you are welcome to start experimenting with. It will have all the features, but we will not have finished the rigorous testing we want to do. I will drop a message in this thread when the release candidate is available.

We will spend a couple more weeks yet testing 3.0.0 to make sure that loading into Snowflake is production ready. When we make the final announcement, we want it to be as robust and as dependable as the Redshift loader.

I am sorry we are missing the February deadline we promised before. I hope you can understand it is because we are targeting a quality first release, instead of rushing it out before it is tested.

2 Likes