This is an automated email from the ASF dual-hosted git repository.
xushiyan pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/hudi.git
The following commit(s) were added to refs/heads/asf-site by this push:
new 1b971e17c47 [DOC] Edit the LSM Timeline blog (#13385)
1b971e17c47 is described below
commit 1b971e17c473cc5a9291c37a34673af1a8b24a36
Author: Dipankar Mazumdar <[email protected]>
AuthorDate: Mon Jun 2 12:34:25 2025 -0400
[DOC] Edit the LSM Timeline blog (#13385)
---
website/blog/2025-05-29-lsm-timeline.md | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/website/blog/2025-05-29-lsm-timeline.md
b/website/blog/2025-05-29-lsm-timeline.md
index e6c97d77e61..6af7496cc32 100644
--- a/website/blog/2025-05-29-lsm-timeline.md
+++ b/website/blog/2025-05-29-lsm-timeline.md
@@ -124,6 +124,12 @@ The LSM timeline unlocks significant advancements in how
Apache Hudi handles met
* **Maintenance-Free Scalability**: The universal compaction and smart
cleaning strategies keep the timeline healthy over time, requiring minimal
manual tuning, while ensuring that old data is cleaned up safely only after
valid snapshots are no longer in use.
+### Performance
+
+Micro-benchmarks show that the LSM Timeline scales efficiently even as the
number of timeline actions grows by orders of magnitude. Reading just the
instant times for `10,000` actions takes around `32ms`, while fetching full
metadata takes `150ms`. At larger scales, such as `10 million` actions,
metadata reads completes in about `162 seconds`.
+
+These results demonstrate that Hudi's LSM timeline can handle decades of
high-frequency commits (e.g., one every 30 seconds for 10+ years) while keeping
metadata access performant.
+
The LSM timeline represents a natural progression in Apache Hudi’s timeline
architecture, designed to address the growing demands of large-scale and
long-lived tables. Hudi’s timeline has been foundational for transactional
integrity, time travel, and incremental processing capabilities. The new
LSM-based design enhances scalability and operational efficiency by introducing
a layered, compacted structure with manifest-driven snapshot isolation. This
improvement allows Hudi to manage exte [...]
---
\ No newline at end of file
