Add batch-llm module: 100,000 synthetic tickets classified by a ChatClient in a Spring Batch 6 chunk step with Resilience4j rate limiting and retry, skip-and-record for malformed JSON, restart after kill -9 in a second JVM, and a per-chunk cost ledger. The model is a deterministic simulated ChatModel; prices are configurable placeholders. No provider called.

Co-Authored-By: Claude Sonnet 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_017ic4b4VQVEJWetTT1dYzyJ
2026-10-29 06:10:22 +00:00
parent aa8b1604ed
commit 6d792f00e0
48 changed files with 2093 additions and 0 deletions
@@ -0,0 +1,80 @@
+package com.ankurm.batchllm;
+
+import java.math.BigDecimal;
+import java.util.List;
+import java.util.concurrent.atomic.AtomicLong;
+
+import org.springframework.batch.infrastructure.item.Chunk;
+import org.springframework.batch.infrastructure.item.ItemWriter;
+import org.springframework.jdbc.core.JdbcTemplate;
+import org.springframework.stereotype.Component;
+import org.springframework.transaction.support.TransactionSynchronization;
+import org.springframework.transaction.support.TransactionSynchronizationManager;
+
+/**
+ * Writes the results AND the cost ledger row for the chunk in one transaction, so a ledger row
+ * exists if and only if its tickets do. The reader position, the results and the ledger either
+ * all commit or all roll back.
+ */
+@Component
+public class ClassificationWriter implements ItemWriter<Classification> {
+
+    private final JdbcTemplate jdbc;
+    private final UsageMeter meter;
+    private final Pricing pricing;
+    private final AtomicLong committedLedgerRows = new AtomicLong();
+
+    public ClassificationWriter(JdbcTemplate jdbc, UsageMeter meter, Pricing pricing) {
+        this.jdbc = jdbc;
+        this.meter = meter;
+        this.pricing = pricing;
+    }
+
+    @Override
+    public void write(Chunk<? extends Classification> chunk) {
+        List<? extends Classification> items = chunk.getItems();
+        jdbc.batchUpdate("INSERT INTO TICKET_RESULT (TICKET_ID, CATEGORY, PRIORITY, CONFIDENCE) VALUES (?,?,?,?)",
+                items, items.size(), (ps, c) -> {
+                    ps.setLong(1, c.ticketId());
+                    ps.setString(2, c.category());
+                    ps.setInt(3, c.priority());
+                    ps.setDouble(4, c.confidence());
+                });
+
+        // A ticket skipped in an earlier, uncommitted attempt (the skip listener commits on its own) and
+        // classified now must not stay in the skipped table.
+        jdbc.batchUpdate("DELETE FROM SKIPPED_TICKET WHERE TICKET_ID = ?", items, items.size(),
+                (ps, c) -> ps.setLong(1, c.ticketId()));
+
+        long first = items.stream().mapToLong(Classification::ticketId).min().orElse(0);
+        long last = items.stream().mapToLong(Classification::ticketId).max().orElse(0);
+        UsageMeter.Snapshot spent = meter.drain();
+        if (TransactionSynchronizationManager.isSynchronizationActive()) {
+            TransactionSynchronizationManager.registerSynchronization(new TransactionSynchronization() {
+                @Override
+                public void afterCompletion(int status) {
+                    if (status == STATUS_COMMITTED) {
+                        long rows = committedLedgerRows.incrementAndGet();
+                        if (rows % 50 == 0) {
+                            System.out.println("PROGRESS committed-chunks-this-jvm=" + rows + " last-ticket=" + last);
+                        }
+                    } else {
+                        meter.restore(spent); // rolled back: the money was still spent
+                    }
+                }
+            });
+        }
+        BigDecimal cost = pricing.cost(spent.promptTokens(), spent.completionTokens());
+        jdbc.update("""
+                INSERT INTO COST_LEDGER (FIRST_TICKET, LAST_TICKET, TICKETS_WRITTEN, TICKETS_SKIPPED, ATTEMPTS,
+                  REJECTED_429, UNAVAILABLE_503, RETRIES, BILLED_CALLS, BAD_REPLIES, PROMPT_TOKENS, COMPLETION_TOKENS,
+                  COST, PRICE_VERSION) VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?,?)""",
+                first, last, items.size(), spent.skipped(), spent.attempts(), spent.rejected429(),
+                spent.unavailable503(), spent.retries(), spent.billedCalls(), spent.badReplies(),
+                spent.promptTokens(), spent.completionTokens(), cost, pricing.version());
+    }
+
+    public long committedLedgerRows() {
+        return committedLedgerRows.get();
+    }
+}