You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@iggy.apache.org by hg...@apache.org on 2025/12/17 15:52:39 UTC
(iggy) 01/02: fix(sdk): update last_flush on empty buffer to prevent busy-loop
This is an automated email from the ASF dual-hosted git repository.
hgruszecki pushed a commit to branch fix/producer-shard-last_flush
in repository https://gitbox.apache.org/repos/asf/iggy.git
commit 33079bb70fce05cd0511479ba9342b768d846ca9
Author: Frank Bell <fr...@eri.rs>
AuthorDate: Mon Dec 8 14:31:54 2025 +0000
fix(sdk): update last_flush on empty buffer to prevent busy-loop
When the linger timeout fires with an empty buffer, last_flush was not
updated. This caused the next deadline to be in the past, making
sleep_until return immediately and creating a tight loop with 100%+ CPU
per shard.
---
core/sdk/src/clients/producer_sharding.rs | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/core/sdk/src/clients/producer_sharding.rs b/core/sdk/src/clients/producer_sharding.rs
index 49c2e17a7..8bdca2213 100644
--- a/core/sdk/src/clients/producer_sharding.rs
+++ b/core/sdk/src/clients/producer_sharding.rs
@@ -164,8 +164,8 @@ impl Shard {
_ = tokio::time::sleep_until(deadline) => {
if !buffer.is_empty() {
Self::flush_buffer(&core, &mut buffer, &mut buffer_bytes, &err_sender).await;
- last_flush = tokio::time::Instant::now();
}
+ last_flush = tokio::time::Instant::now();
}
_ = stop_rx.recv() => {
closed_clone.store(true, Ordering::Release);