Create table concurrently with subscribing to stream of events #77
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
For KCL apps, we want the KCL to get initialized as early as possible, so the worker can claim shard leases before they get stolen by other workers. Before this PR, the loader initialized the destination table first, and then subscribed to the stream afterwards. Initializing the destination table can be fairly slow, especially because we do things like syncing to the external catalog, and possibly cleaning up aborted commits. After this PR, the loader subscribes to the stream concurrently with initializing the destination table. This lets the KCL claim leases before they get stolen.
pondzix
approved these changes
Aug 8, 2024
zhaow-de
added a commit
to alloy-ch/rcplus-alloy-snowplow-lake-loader
that referenced
this pull request
Oct 4, 2024
…patch-for-alloy * commit '7ab2edc3fd4d81ffb4d5f3285d02330def7672b1': Upgrade common-streams to 0.8.0-M5 Delete files asynchronously (snowplow-incubator#82) Upgrade common-streams 0.8.0-M4 (snowplow-incubator#81) Avoid error on duplicate view name (snowplow-incubator#80) Add option to exit on missing Iglu schemas (snowplow-incubator#79) common-streams 0.8.x with refactored health monitoring (snowplow-incubator#78) Create table concurrently with subscribing to stream of events (snowplow-incubator#77) Iceberg fail fast if missing permissions on the catalog (snowplow-incubator#76) Make alert messages more human-readable (snowplow-incubator#75) Hudi loader should fail early if missing permissions on Glue catalog (snowplow-incubator#72) Add alert & retry for delta/s3 initialization (snowplow-incubator#74) Implement alerting and retrying mechanisms Bump aws-hudi to 1.0.0-beta2 (snowplow-incubator#71) Bump hudi to 0.15.0 (snowplow-incubator#70) Allow disregarding Iglu field's nullability when creating output columns (snowplow-incubator#66) Extend health probe to report unhealthy on more error scenarios (snowplow-incubator#69) Fix bad rows resizing (snowplow-incubator#68)
oguzhanunlu
pushed a commit
that referenced
this pull request
Nov 1, 2024
For KCL apps, we want the KCL to get initialized as early as possible, so the worker can claim shard leases before they get stolen by other workers. Before this PR, the loader initialized the destination table first, and then subscribed to the stream afterwards. Initializing the destination table can be fairly slow, especially because we do things like syncing to the external catalog, and possibly cleaning up aborted commits. After this PR, the loader subscribes to the stream concurrently with initializing the destination table. This lets the KCL claim leases before they get stolen.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
For KCL apps, we want the KCL to get initialized as early as possible, so the worker can claim shard leases before they get stolen by other workers.
Before this PR, the loader initialized the destination table first, and then subscribed to the stream afterwards. Initializing the destination table can be fairly slow, especially because we do things like syncing to the external catalog, and possibly cleaning up aborted commits.
After this PR, the loader subscribes to the stream concurrently with initializing the destination table. This lets the KCL claim leases before they get stolen.