[ 
https://issues.apache.org/jira/browse/KNOX-2990?focusedWorklogId=902905&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-902905
 ]

ASF GitHub Bot logged work on KNOX-2990:
----------------------------------------

                Author: ASF GitHub Bot
            Created on: 31/Jan/24 21:21
            Start Date: 31/Jan/24 21:21
    Worklog Time Spent: 10m 
      Work Description: pzampino commented on code in PR #826:
URL: https://github.com/apache/knox/pull/826#discussion_r1473463201


##########
gateway-server/src/main/java/org/apache/knox/gateway/services/factory/TokenStateServiceFactory.java:
##########
@@ -61,17 +62,32 @@ protected Service createService(GatewayServices 
gatewayServices, ServiceType ser
           service.init(gatewayConfig, options);
         } catch (ServiceLifecycleException e) {
           LOG.errorInitializingService(implementation, e.getMessage(), e);
-          service = new AliasBasedTokenStateService();
-          ((AliasBasedTokenStateService) 
service).setAliasService(getAliasService(gatewayServices));
+          service = useDerbyDatabaseTokenStateService(gatewayServices, 
gatewayConfig, options);
         }
+      } else if (matchesImplementation(implementation, 
DerbyDBTokenStateService.class, true)) {
+        service = useDerbyDatabaseTokenStateService(gatewayServices, 
gatewayConfig, options);
       }
 
-      logServiceUsage(isEmptyDefaultImplementation(implementation) ? 
AliasBasedTokenStateService.class.getName() : implementation, serviceType);
+      logServiceUsage(service.getClass().getName(), serviceType);
     }
 
     return service;
   }
 
+  private Service useDerbyDatabaseTokenStateService(GatewayServices 
gatewayServices, GatewayConfig gatewayConfig, Map<String, String> options) {
+    Service service;
+    try {
+      service = new DerbyDBTokenStateService();
+      ((DerbyDBTokenStateService) 
service).setAliasService(getAliasService(gatewayServices));

Review Comment:
   If the constructors accept GatewayServices, GatewayConfig, and options, the 
implementations can do all their own initialization in a self-contained manner, 
simplifying the factory implementation.



##########
gateway-server/src/main/java/org/apache/knox/gateway/services/token/impl/JDBCTokenStateService.java:
##########
@@ -65,12 +76,33 @@ public void init(GatewayConfig config, Map<String, String> 
options) throws Servi
         } catch (Exception e) {
           throw new ServiceLifecycleException("Error while initiating 
JDBCTokenStateService: " + e, e);
         }
+
+        this.skipTokenMigration = config.skipTokenMigration();
+        this.archiveMigratedTokens = config.archiveMigratedTokens();
+        this.migrateExpiredTokens = config.migrateExpiredTokens();
+        this.verboseTokenMigration = 
config.printVerboseTokenMigrationMessages();
+        this.tokenMigrationProgressCount = 
config.getTokenMigrationProgressCount();
       } finally {
         initLock.unlock();
       }
     }
   }
 
+  @Override
+  public void start() throws ServiceLifecycleException {
+    super.start();
+    if (skipTokenMigration) {
+      log.skipTokenMigration();
+    } else {
+      final TokenMigrationTool tokenMigrationTool = new 
TokenMigrationTool(aliasService, this, null);

Review Comment:
   Token migration is an implicit action when a JDBC-based impl starts? I 
missed that part when I first read the PR description. Should 
skipTokenMigration default to true, such that an admin has to explicitly enable 
this implicit migration?



##########
gateway-server/src/main/java/org/apache/knox/gateway/services/token/impl/DefaultTokenStateService.java:
##########
@@ -457,4 +457,9 @@ private Collection<KnoxToken> fetchTokens(String userName, 
boolean createdBy) {
     });
     return tokens;
   }
+
+  @Override
+  public boolean isMigrationTarget() {

Review Comment:
   Instead of a method to override in every impl, you could have those 
implementations that support migration implement a MigrationTarget interface.



##########
gateway-server/src/main/java/org/apache/knox/gateway/services/factory/TokenStateServiceFactory.java:
##########
@@ -61,17 +62,32 @@ protected Service createService(GatewayServices 
gatewayServices, ServiceType ser
           service.init(gatewayConfig, options);
         } catch (ServiceLifecycleException e) {
           LOG.errorInitializingService(implementation, e.getMessage(), e);
-          service = new AliasBasedTokenStateService();
-          ((AliasBasedTokenStateService) 
service).setAliasService(getAliasService(gatewayServices));
+          service = useDerbyDatabaseTokenStateService(gatewayServices, 
gatewayConfig, options);
         }
+      } else if (matchesImplementation(implementation, 
DerbyDBTokenStateService.class, true)) {

Review Comment:
   There is something about this logic that smells bad to me. From my point of 
view, someone should be able to implement TokenStateService with whatever impl 
they need, configure it in gateway-site, and drop their JAR in the ext dir. 
With this logic, that won't work. Can we consider refactoring this to keep Knox 
extensible?



##########
gateway-server/src/main/java/org/apache/knox/gateway/services/factory/TokenStateServiceFactory.java:
##########
@@ -61,17 +62,32 @@ protected Service createService(GatewayServices 
gatewayServices, ServiceType ser
           service.init(gatewayConfig, options);
         } catch (ServiceLifecycleException e) {
           LOG.errorInitializingService(implementation, e.getMessage(), e);
-          service = new AliasBasedTokenStateService();
-          ((AliasBasedTokenStateService) 
service).setAliasService(getAliasService(gatewayServices));
+          service = useDerbyDatabaseTokenStateService(gatewayServices, 
gatewayConfig, options);
         }
+      } else if (matchesImplementation(implementation, 
DerbyDBTokenStateService.class, true)) {
+        service = useDerbyDatabaseTokenStateService(gatewayServices, 
gatewayConfig, options);
       }
 
-      logServiceUsage(isEmptyDefaultImplementation(implementation) ? 
AliasBasedTokenStateService.class.getName() : implementation, serviceType);
+      logServiceUsage(service.getClass().getName(), serviceType);
     }
 
     return service;
   }
 
+  private Service useDerbyDatabaseTokenStateService(GatewayServices 
gatewayServices, GatewayConfig gatewayConfig, Map<String, String> options) {
+    Service service;
+    try {
+      service = new DerbyDBTokenStateService();
+      ((DerbyDBTokenStateService) 
service).setAliasService(getAliasService(gatewayServices));
+      ((DerbyDBTokenStateService) 
service).setMasterService(getMasterService(gatewayServices));
+      service.init(gatewayConfig, options);
+    } catch (ServiceLifecycleException e) {
+      LOG.errorInitializingService(DerbyDBTokenStateService.class.getName(), 
e.getMessage(), e);
+      service = new DefaultTokenStateService();

Review Comment:
   Falling back to in-memory is ok with me, so long as we log some indication 
about that fact. If a user thinks their Knox is running against the DB they 
configured, but it's only in-memory, that could be a real problem.



##########
gateway-server/src/main/java/org/apache/knox/gateway/util/TokenMigrationTool.java:
##########
@@ -0,0 +1,232 @@
+/*
+ * 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.knox.gateway.util;
+
+import java.io.PrintStream;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.Locale;
+import java.util.Map;
+import java.util.Set;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.atomic.AtomicInteger;
+
+import org.apache.knox.gateway.i18n.messages.MessagesFactory;
+import org.apache.knox.gateway.services.security.AliasService;
+import org.apache.knox.gateway.services.security.AliasServiceException;
+import org.apache.knox.gateway.services.security.token.TokenMetadata;
+import org.apache.knox.gateway.services.security.token.TokenStateService;
+import org.apache.knox.gateway.services.token.impl.TokenStateServiceMessages;
+
+public class TokenMigrationTool {
+
+  private static final String TOKEN_ALIAS_SUFFIX_DELIM = "--";
+  private static final String TOKEN_ISSUE_TIME_POSTFIX = 
TOKEN_ALIAS_SUFFIX_DELIM + "iss";

Review Comment:
   It would be ideal if these were referenced from the 
AliasBasedTokenStateService where they're defined.





Issue Time Tracking
-------------------

    Worklog Id:     (was: 902905)
    Time Spent: 1.5h  (was: 1h 20m)

> TokenStateService implementation cleanup
> ----------------------------------------
>
>                 Key: KNOX-2990
>                 URL: https://issues.apache.org/jira/browse/KNOX-2990
>             Project: Apache Knox
>          Issue Type: Task
>          Components: Server
>    Affects Versions: 2.0.0, 1.6.0, 1.6.1
>            Reporter: Sandor Molnar
>            Assignee: Sandor Molnar
>            Priority: Critical
>             Fix For: 2.1.0
>
>          Time Spent: 1.5h
>  Remaining Estimate: 0h
>
> This issue is driven by a [DISCUSS] thread initiated on Knox's DEV mailing 
> list [here|https://lists.apache.org/thread/fs9nkl6l45o330ttvgvqxj3jnxt63bcs].
> As a result of that discussion, the following needs to be implemented:
>  * deprecate the following TSS implementations:
>  ** AliasBasedTokenStateService
>  ** ZookeeperTokenStateService
>  ** JournalBasedTokenStateService
>  * document the deprecation of these TSS implementations in v2.1.0 and 
> highlight that they will be removed in the upcoming release (v2.2.0?).
>  * implement a DerbyDB storage that will store tokens in 
> {{$DATA_DIR/security/tokens}} (encrypted or not, it'll be decided later)
>  * make sure appropriate file permissions are set on that folder
>  * have the {{homepage}} topology configured with JDBC TSS pointing to this 
> DerbyDB storage
>  * implement a new KnoxCLI command that migrates existing tokens from 
> credential stores to the DerbyDB storage
>  * automate this new KnoxCLI command in a way such that it runs when Knox 
> Gateway is started, token management is enabled, and DerbyDB storage is 
> configured
>  * ensure that the previous automated step can be controlled (E.g. in case of 
> unforeseen errors it can be turned off)
>  * document possible data replication scenarios when, in the case of HA 
> deployments, existing tokens from one Knox node should be made available in 
> other Knox node(s) and there is no other centralized RDBMS in use 
> (PostgreSQL, MySQL for instance)
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to