From 6ae39b4ea52a1aa5da04dc04530605639e345f7d Mon Sep 17 00:00:00 2001
From: trifonovt <87468028+TihomirTrifonov@users.noreply.github.com>
Date: Thu, 9 Apr 2026 12:53:11 +0200
Subject: [PATCH] batch embedding support

---
 src/main/resources/application-new.yml | 32 ++++++++++++++++++++++----
 1 file changed, 28 insertions(+), 4 deletions(-)

diff --git a/src/main/resources/application-new.yml b/src/main/resources/application-new.yml
index 4feb0bf..1bb47af 100644
--- a/src/main/resources/application-new.yml
+++ b/src/main/resources/application-new.yml
@@ -1,6 +1,7 @@
 dip:
   runtime:
     mode: NEW
+
   search:
       # Default page size for search results
       default-page-size: 20
@@ -36,9 +37,11 @@ dip:
       debug-top-hits-per-engine: 10
 
   embedding:
-      enabled: true
+      enabled: false
       jobs:
         enabled: true
+        process-in-batches: true
+        execution-batch-size: 20
 
       default-document-model: e5-default
       default-query-model: e5-default
@@ -58,10 +61,14 @@ dip:
         vector-sync-e5:
           type: http-vector-sync
           base-url: http://localhost:8001
-          connect-timeout: 5s
-          read-timeout: 60s
+          connect-timeout: 30s
+          read-timeout: 300s
           headers:
             X-Client: dip
+          batch-request:
+            truncate-text: true
+            truncate-length: 512
+            chunk-size: 8
 
       models:
 
@@ -79,6 +86,7 @@ dip:
           dimensions: 1024
           distance-metric: COSINE
           supports-query-embedding-mode: true
+          supports-batch: true
           active: true
 
       profiles:
@@ -215,7 +223,7 @@ dip:
     # ted packages download configuration
     ted-download:
       # Enable/disable automatic package download
-      enabled: true
+      enabled: false
       # Base URL for TED Daily Packages
       base-url: https://ted.europa.eu/packages/daily/
       # Download directory for tar.gz files
@@ -238,6 +246,22 @@ dip:
       delay-between-downloads: 5000
       # Delete tar.gz after ingestion
       delete-after-ingestion: true
+
+  migration:
+    legacy-audit:
+      # Enable/disable the Wave 1 / Milestone A legacy integrity audit subsystem
+      enabled: true
+      # Optional startup execution; the audit is read-only and only writes audit run/finding tables
+      startup-run-enabled: true
+      # Maximum number of legacy TED documents to scan during startup (0 = all)
+      startup-run-limit: 0
+      # Page size for legacy TED document scanning
+      page-size: 100
+      # Maximum number of persisted findings in a single run
+      max-findings-per-run: 10000
+      # Maximum number of grouped duplicate samples captured for aggregate checks
+      max-duplicate-samples: 100
+
   ted: # Phase 3 TED projection configuration
     projection:
       # Enable/disable dual-write into the TED projection model on top of DOC.doc_document