From 42d5d900a9e39dc81bafca3801e8c3f4edc050fd Mon Sep 17 00:00:00 2001 From: Chris Lu Date: Sun, 15 Mar 2026 10:03:15 -0700 Subject: [PATCH] Update Iceberg-Table-Maintenance.md --- Iceberg-Table-Maintenance.md | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/Iceberg-Table-Maintenance.md b/Iceberg-Table-Maintenance.md index 0c72f28..6365556 100644 --- a/Iceberg-Table-Maintenance.md +++ b/Iceberg-Table-Maintenance.md @@ -273,7 +273,7 @@ Result: `"rewrote 12 manifests into 1 (320 entries)"` | Condition | Result | |:---|:---| -| Manifest count < `min_input_files` | `"only N manifests, below threshold"` | +| Manifest count < `min_manifests_to_rewrite` | `"only N manifests, below threshold of M"` | | No current snapshot | `"no current snapshot"` | | No data entries | `"no data entries to rewrite"` | @@ -307,7 +307,8 @@ These control **how** maintenance operations behave. Set via the Admin UI or pas | `max_snapshots_to_keep` | `5` | Always keep at least this many newest snapshots | | `orphan_older_than_hours` | `72` (3 days) | Safety window: only delete orphans older than this | | `target_file_size_bytes` | `268435456` (256 MB) | Files smaller than this are compaction candidates | -| `min_input_files` | `5` | Minimum files in a partition/manifest group to trigger action | +| `min_input_files` | `5` | Minimum small files in a partition to trigger compaction | +| `min_manifests_to_rewrite` | `5` | Minimum manifests before rewriting is triggered | | `max_commit_retries` | `5` | Max optimistic concurrency retries on version conflict | | `operations` | `all` | Comma-separated list of operations, or `all` | @@ -400,7 +401,7 @@ rewrite_manifests: rewrote 7 manifests into 1 (120 entries) - **Client-side CAS**: the compare-and-swap on metadata version is enforced client-side. Concurrent maintenance on the same table should be avoided (the deduplication key prevents this under normal scheduling). - **No delete manifest support in compaction**: tables with row-level deletes (delete manifests) skip compaction. - **No multi-spec compaction**: tables that have undergone partition spec evolution skip compaction. -- **Single filer**: detection and execution connect to one filer address from the cluster context. +- **Single filer per job**: detection tries each filer address in the cluster context until one connects, but each execution job uses the single filer address recorded in its proposal. ## See Also