Pace Turbo pipelines with the throttle transform
Follow @goldskyio on Twitter to stay updated on everything we ship.
The new throttle transform caps how fast data moves through a Turbo pipeline. Use it to stay under a downstream API's rate limit or smooth out a bursty source.
Throttle groups records into batches and emits one batch per interval. It doesn't change your data, only when it arrives.
transforms:
my_throttle:
type: throttle
from: my_source
max_batch_size: 100
min_batch_interval: 10s
Bounded backfills on Edge EVM datasets
Edge EVM datasets now support end_at. Set it alongside start_at to backfill a fixed block range.
- Added chain support: Gensyn Mainnet (Subgraphs)
- Turbo sources, transforms, and sinks now accept
batch_sizeandbatch_interval. S3 sinks benefit most - ClickHouse sinks now surface table creation errors, like a nullable column in the primary key, so you can fix the config right away
- The Active tab on the Pipelines page now includes Turbo pipelines
- New Compose example apps: a prediction market and a proof of reserves app
- Compose app templates now ship with
AGENT.mdandSKILLS.mdfiles for AI coding assistants - Compose now redeploys the Turbo pipeline behind an onchain trigger only when the trigger itself changes
- Fixed duplicate Compose failure emails
- Fixed charts not resizing when you collapse the details pane