yadavay-amzn commented on code in PR #57262:
URL: https://github.com/apache/spark/pull/57262#discussion_r3583874029


##########
core/src/main/java/org/apache/spark/security/credentials/FileTokenIngestor.java:
##########
@@ -0,0 +1,139 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.security.credentials;
+
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.time.Instant;
+import java.util.Date;
+import java.util.Optional;
+
+import io.jsonwebtoken.Claims;
+import io.jsonwebtoken.Jwts;
+
+import org.apache.spark.annotation.DeveloperApi;
+import org.apache.spark.internal.SparkLogger;
+import org.apache.spark.internal.SparkLoggerFactory;
+import org.apache.spark.security.UserContext;
+
+/**
+ * :: DeveloperApi ::
+ * A {@link TokenIngestor} that reads an OIDC identity token from a file.
+ * <p>
+ * The file path is typically a Kubernetes projected service account token
+ * (e.g., {@code /var/run/secrets/tokens/spark-identity}) or a path configured 
via
+ * {@code spark.security.credentials.identityToken.file}.
+ * <p>
+ * This implementation:
+ * <ul>
+ *   <li>Detects file rotation via mtime change (only re-parses when the file 
changes)</li>
+ *   <li>Parses JWT claims without signature verification (unsigned claims 
parsing)
+ *       since the token is trusted from the local filesystem</li>
+ *   <li>Handles errors gracefully: malformed JWT, missing file, or empty 
content
+ *       returns empty rather than throwing</li>
+ * </ul>
+ *
+ * @since 4.3.0
+ */
+@DeveloperApi
+public class FileTokenIngestor implements TokenIngestor {
+
+  private static final SparkLogger LOG = 
SparkLoggerFactory.getLogger(FileTokenIngestor.class);
+
+  private final Path tokenPath;
+  private volatile long lastMtime = -1L;
+  private volatile UserContext cachedContext = null;
+
+  /**
+   * Construct a new FileTokenIngestor.
+   *
+   * @param tokenPath path to the identity token file (must not be null)
+   */
+  public FileTokenIngestor(Path tokenPath) {
+    this.tokenPath = tokenPath;
+  }
+
+  @Override
+  public Optional<UserContext> load() {
+    try {
+      if (!Files.exists(tokenPath)) {
+        LOG.debug("Token file does not exist: {}", tokenPath);
+        return Optional.empty();
+      }
+
+      // File did not change since last successful parse
+      long currentMtime = Files.getLastModifiedTime(tokenPath).toMillis();
+      if (currentMtime == lastMtime && cachedContext != null) {
+        return Optional.of(cachedContext);
+      }
+
+      // Read the token contents and check for emptiness
+      String content = new String(Files.readAllBytes(tokenPath), 
"UTF-8").trim();
+      if (content.isEmpty()) {
+        LOG.warn("Token file is empty: " + tokenPath);
+        return Optional.empty();
+      }
+
+      Optional<UserContext> userContext = parseJwt(content);
+      if (userContext.isPresent()) {
+        lastMtime = currentMtime;
+        cachedContext = userContext.get();
+      }
+      return userContext;
+    } catch (Exception e) {
+      LOG.warn("Failed to load token from " + tokenPath, e);
+      return Optional.empty();
+    }
+  }
+
+  /**
+   * Parse a JWT token string into a UserContext using unsigned claims parsing.
+   * The token is trusted from the local filesystem. No need to verify the 
signature.
+   */
+  private Optional<UserContext> parseJwt(String token) {
+    try {
+      Claims claims = Jwts.parser()
+          .unsecured()
+          .build()
+          .parseUnsecuredClaims(token)

Review Comment:
   I think there might be an issue here: this only accepts unsigned JWTs (`alg: 
none`), but a Kubernetes projected service account token is a signed JWT 
(RS256), as is any real OIDC IdP token. I tried it against jjwt 0.13.0 and a 
signed token throws `UnsupportedJwtException: Cannot verify JWS signature: 
unable to locate signature verification key`, so it looks like `load()` would 
return empty for the production input this ingestor targets. Stripping the 
signature and passing `header.payload.` didn't work for me either, since jjwt 
sees the `alg` header and still requires the signature.
   
   Since the token is trusted (local FS, kubelet-delivered) and the signature 
is re-verified downstream at exchange, I think a possible fix is to read the 
claims without going through the JWS parser, by decoding the payload segment 
directly rather than routing a signed token through `parseUnsecuredClaims`. I 
confirmed `Base64.getUrlDecoder().decode(parts[1])` gives back the claims JSON 
for a signed token.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to