Github user hanm commented on a diff in the pull request:
https://github.com/apache/zookeeper/pull/338#discussion_r138181648
--- Diff: src/java/main/org/apache/zookeeper/audit/AuditConstants.java ---
@@ -0,0 +1,37 @@
+/**
+ * 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.zookeeper.audit;
+
+public class AuditConstants {
+ public static final String SUCCESS = "success";
+ public static final String FAILURE = "failure";
+ // operation is performed, result is not known yet
+ public static final String INVOKED = "invoked";
+ public static final String KEY_VAL_SEPARATOR = "=";
+ public static final char PAIR_SEPARATOR = '\t';
+
+ public static final String OP_START = "serverStart";
--- End diff --
I think we should not expose the internal op codes, etc to end users,
through API or through the public constant definition included in this patch.
We can get rid of the constants definitions here and instead, whenever we want
to audit log something we create the operation string on site and directly pass
the string to the audit event function call. Something like:
`case OpCode.create:
subResult = new CreateResult(subTxnResult.path);
addSuccessAudit(request, cnxn, "create", subTxnResult.path);
break;
`
This approach seems make the maintain overhead low as we don't have to
worry about the correlations between various op codes definitions and the only
catch is we might have to duplicate the op code string in different places if
we audit logging same thing in different places - but in this case it seems the
cost of duplication is low comparing to maintaining explicit op constants.
I also checked HDFS code base and it took a similar approach.
---
