[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320588988
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
+ LOG.error(msg);
+ throw new SCMException(msg, SCMException.ResultCodes
+ .FAILED_TO_FIND_HEALTHY_NODES);
+}
+
+if (healthyNodes.size() < nodesRequired) {
+ msg = String.format("Not enough healthy nodes to allocate container. %d "
+ + " datanodes required. Found %d",
+ nodesRequired, healthyNodes.size());
+ LOG.er
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320588221
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
+ LOG.error(msg);
+ throw new SCMException(msg, SCMException.ResultCodes
+ .FAILED_TO_FIND_HEALTHY_NODES);
+}
+
+if (healthyNodes.size() < nodesRequired) {
+ msg = String.format("Not enough healthy nodes to allocate container. %d "
+ + " datanodes required. Found %d",
+ nodesRequired, healthyNodes.size());
+ LOG.er
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320559908
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
+ LOG.error(msg);
+ throw new SCMException(msg, SCMException.ResultCodes
+ .FAILED_TO_FIND_HEALTHY_NODES);
+}
+
+if (healthyNodes.size() < nodesRequired) {
+ msg = String.format("Not enough healthy nodes to allocate container. %d "
+ + " datanodes required. Found %d",
+ nodesRequired, healthyNodes.size());
+ LOG.er
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320557920
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
Review comment:
Improve these comments
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
-
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320555713
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
+ LOG.error(msg);
+ throw new SCMException(msg, SCMException.ResultCodes
+ .FAILED_TO_FIND_HEALTHY_NODES);
+}
+
+if (healthyNodes.size() < nodesRequired) {
+ msg = String.format("Not enough healthy nodes to allocate container. %d "
+ + " datanodes required. Found %d",
+ nodesRequired, healthyNodes.size());
+ LOG.er
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320555432
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
+ LOG.error(msg);
+ throw new SCMException(msg, SCMException.ResultCodes
+ .FAILED_TO_FIND_HEALTHY_NODES);
+}
+
+if (healthyNodes.size() < nodesRequired) {
+ msg = String.format("Not enough healthy nodes to allocate container. %d "
Review comment:
container -> pipeline
T
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default
pipeline placement policy implementation.
URL: https://github.com/apache/hadoop/pull/1366#discussion_r320555339
##
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/pipeline/PipelinePlacementPolicy.java
##
@@ -0,0 +1,291 @@
+/**
+ * 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.hadoop.hdds.scm.pipeline;
+
+import com.google.common.annotations.VisibleForTesting;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdds.protocol.DatanodeDetails;
+import org.apache.hadoop.hdds.protocol.proto.HddsProtos;
+import org.apache.hadoop.hdds.scm.ScmConfigKeys;
+import
org.apache.hadoop.hdds.scm.container.placement.algorithms.SCMCommonPolicy;
+import org.apache.hadoop.hdds.scm.container.placement.metrics.SCMNodeMetric;
+import org.apache.hadoop.hdds.scm.exceptions.SCMException;
+import org.apache.hadoop.hdds.scm.net.NetworkTopology;
+import org.apache.hadoop.hdds.scm.net.Node;
+import org.apache.hadoop.hdds.scm.node.NodeManager;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Pipeline placement policy that choose datanodes based on load balancing
+ * and network topology to supply pipeline creation.
+ *
+ * 1. get a list of healthy nodes
+ * 2. filter out nodes that are not too heavily engaged in other pipelines
+ * 3. Choose an anchor node among the viable nodes which follows the algorithm
+ * described @SCMContainerPlacementCapacity
+ * 4. Choose other nodes around the anchor node based on network topology
+ */
+public final class PipelinePlacementPolicy extends SCMCommonPolicy {
+ @VisibleForTesting
+ static final Logger LOG =
+ LoggerFactory.getLogger(PipelinePlacementPolicy.class);
+ private final NodeManager nodeManager;
+ private final Configuration conf;
+ private final int heavyNodeCriteria;
+
+ /**
+ * Constructs a Container Placement with considering only capacity.
+ * That is this policy tries to place containers based on node weight.
+ *
+ * @param nodeManager Node Manager
+ * @param confConfiguration
+ */
+ public PipelinePlacementPolicy(final NodeManager nodeManager,
+ final Configuration conf) {
+super(nodeManager, conf);
+this.nodeManager = nodeManager;
+this.conf = conf;
+heavyNodeCriteria = conf.getInt(
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT,
+ScmConfigKeys.OZONE_DATANODE_MAX_PIPELINE_ENGAGEMENT_DEFAULT);
+ }
+
+ /**
+ * Returns true if this node meets the criteria.
+ *
+ * @param datanodeDetails DatanodeDetails
+ * @return true if we have enough space.
+ */
+ @VisibleForTesting
+ boolean meetCriteria(DatanodeDetails datanodeDetails,
+ long heavyNodeLimit) {
+return (nodeManager.getPipelinesCount(datanodeDetails) <= heavyNodeLimit);
+ }
+
+ /**
+ * Filter out viable nodes based on
+ * 1. nodes that are healthy
+ * 2. nodes that are not too heavily engaged in other pipelines
+ *
+ * @param excludedNodes - excluded nodes
+ * @param nodesRequired - number of datanodes required.
+ * @return a list of viable nodes
+ * @throws SCMException when viable nodes are not enough in numbers
+ */
+ List filterViableNodes(
+ List excludedNodes, int nodesRequired)
+ throws SCMException {
+// get nodes in HEALTHY state
+List healthyNodes =
+nodeManager.getNodes(HddsProtos.NodeState.HEALTHY);
+if (excludedNodes != null) {
+ healthyNodes.removeAll(excludedNodes);
+}
+String msg;
+if (healthyNodes.size() == 0) {
+ msg = "No healthy node found to allocate container.";
Review comment:
"container" -> pipeline
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With reg
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation. URL: https://github.com/apache/hadoop/pull/1366#discussion_r320554141 ## File path: hadoop-ozone/pom.xml ## @@ -19,7 +19,7 @@ 3.2.0 - hadoop-ozone + hadoop-_ozone Review comment: What's this for? 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation. URL: https://github.com/apache/hadoop/pull/1366#discussion_r320553505 ## File path: hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/node/NodeManager.java ## @@ -199,4 +220,10 @@ void processNodeReport(DatanodeDetails datanodeDetails, * @return the given datanode, or null if not found */ DatanodeDetails getNodeByAddress(String address); + + /** + * Get cluster map as in network topology for this node manager. + * @return cluster map + */ + NetworkTopology getClusterMap(); Review comment: Suggest to change the name to getClusterNetworkTopologyMap. 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[GitHub] [hadoop] ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation.
ChenSammi commented on a change in pull request #1366: HDDS-1577. Add default pipeline placement policy implementation. URL: https://github.com/apache/hadoop/pull/1366#discussion_r320552958 ## File path: hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/node/NodeManager.java ## @@ -129,6 +138,18 @@ */ void removePipeline(Pipeline pipeline); + /** + * Get the entire Node2PipelineMap. + * @return Node2PipelineMap + */ + Node2PipelineMap getNode2PipelineMap(); + + /** + * Set the Node2PipelineMap. + * @param node2PipelineMap Node2PipelineMap + */ + void setNode2PipelineMap(Node2PipelineMap node2PipelineMap); Review comment: I would suggest to remove this function from interface definition since it's only used in test. 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
