You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by GitBox <gi...@apache.org> on 2020/11/18 14:01:14 UTC
[GitHub] [lucene-solr] shalinmangar opened a new pull request #2086: SOLR-14985: Slow indexing and search performance when using HttpClusterStateProvider
shalinmangar opened a new pull request #2086:
URL: https://github.com/apache/lucene-solr/pull/2086
# Description
HttpClusterStateProvider fetches and caches Aliases and Live Nodes for 5 seconds. But all calls to getState are live. These collection states are supposed to be cached inside BaseSolrCloudClient but BaseSolrCloudClient only caches collection state if it is lazy which is never the case for states returned by HttpClusterStateProvider.
The BaseSolrCloudClient calls getState for each collection mentioned in the request thereby making a live http call. It also calls getClusterProperties() for each live node!
So overall, at least 4 HTTP calls are made to fetch cluster state for each update request when using HttpClusterStateProvider. There may be more if aliases are involved or if more than one collection is specified in the request. Similar problems exist on the query path as well.
Due to these reasons, using HttpClusterStateProvider causes horrible latencies and throughput for update and search requests.
# Solution
This PR fixes BaseCloudSolrClient to cache collection states returned by HttpClusterStateProvider, reduce the number of calls to getClusterProperty in case of admin requests, replace usage of getClusterStateProvider().getState() with getDocCollection() which caches the collection state. Therefore the number of clusterstatus calls are reduced from 4 for each query/indexing to either one at max and usually 0 (if data is cached already).
# Tests
A new CountingHttpClusterStateProvider test class is added which can track the number of http calls made. This class is used in the HttpClusterStateProviderTest to track and assert the number of http calls made by HttpClusterStateProvider.
# Checklist
Please review the following and check all that apply:
- [X] I have reviewed the guidelines for [How to Contribute](https://wiki.apache.org/solr/HowToContribute) and my code conforms to the standards described there to the best of my ability.
- [X] I have created a Jira issue and added the issue ID to my pull request title.
- [X] I have developed this patch against the `master` branch.
- [X] I have run `./gradlew check`.
- [X] I have added tests for my changes.
- [ ] I have added documentation for the [Ref Guide](https://github.com/apache/lucene-solr/tree/master/solr/solr-ref-guide) (for Solr changes only).
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene-solr] madrob commented on pull request #2086: SOLR-14985: Slow indexing and search performance when using HttpClusterStateProvider
Posted by GitBox <gi...@apache.org>.
madrob commented on pull request #2086:
URL: https://github.com/apache/lucene-solr/pull/2086#issuecomment-729873570
@thelabdude this seems like something you'd be interested in was well.
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene-solr] madrob commented on a change in pull request #2086: SOLR-14985: Slow indexing and search performance when using HttpClusterStateProvider
Posted by GitBox <gi...@apache.org>.
madrob commented on a change in pull request #2086:
URL: https://github.com/apache/lucene-solr/pull/2086#discussion_r526311005
##########
File path: solr/solrj/src/test/org/apache/solr/client/solrj/impl/CountingHttpClusterStateProvider.java
##########
@@ -0,0 +1,72 @@
+/*
+ * 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.solr.client.solrj.impl;
+
+import org.apache.http.client.HttpClient;
+import org.apache.solr.client.solrj.ResponseParser;
+import org.apache.solr.client.solrj.SolrClient;
+import org.apache.solr.client.solrj.SolrRequest;
+import org.apache.solr.client.solrj.SolrServerException;
+import org.apache.solr.common.util.NamedList;
+
+import java.io.IOException;
+import java.util.List;
+import java.util.concurrent.atomic.AtomicInteger;
+
+@SuppressWarnings({"unchecked"})
+public class CountingHttpClusterStateProvider extends BaseHttpClusterStateProvider {
+
+ private final HttpClient httpClient;
+ private final boolean clientIsInternal;
+
+ private final AtomicInteger counter = new AtomicInteger(0);
+
+ public CountingHttpClusterStateProvider(List<String> solrUrls, HttpClient httpClient) throws Exception {
+ this.httpClient = httpClient == null ? HttpClientUtil.createClient(null) : httpClient;
+ this.clientIsInternal = httpClient == null;
+ init(solrUrls);
+ }
+
+ @Override
+ protected SolrClient getSolrClient(String baseUrl) {
+ return new AssertingHttpSolrClient(new HttpSolrClient.Builder().withBaseSolrUrl(baseUrl).withHttpClient(httpClient));
+ }
+
+ @Override
+ public void close() throws IOException {
+ if (this.clientIsInternal && this.httpClient != null) {
+ HttpClientUtil.close(httpClient);
+ }
+ }
+
+ public int getRequestCount() {
+ return counter.get();
+ }
+
+ class AssertingHttpSolrClient extends HttpSolrClient {
+ public AssertingHttpSolrClient(Builder builder) {
+ super(builder);
+ }
+
+ @Override
+ public NamedList<Object> request(@SuppressWarnings({"rawtypes"}) SolrRequest request, ResponseParser processor, String collection) throws SolrServerException, IOException {
+ new Exception().printStackTrace();
Review comment:
please send this to a logger.
##########
File path: solr/solrj/src/test/org/apache/solr/client/solrj/impl/CountingHttpClusterStateProvider.java
##########
@@ -0,0 +1,72 @@
+/*
+ * 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.solr.client.solrj.impl;
+
+import org.apache.http.client.HttpClient;
+import org.apache.solr.client.solrj.ResponseParser;
+import org.apache.solr.client.solrj.SolrClient;
+import org.apache.solr.client.solrj.SolrRequest;
+import org.apache.solr.client.solrj.SolrServerException;
+import org.apache.solr.common.util.NamedList;
+
+import java.io.IOException;
+import java.util.List;
+import java.util.concurrent.atomic.AtomicInteger;
+
+@SuppressWarnings({"unchecked"})
Review comment:
unneeded
##########
File path: solr/solrj/src/test/org/apache/solr/client/solrj/impl/HttpClusterStateProviderTest.java
##########
@@ -0,0 +1,95 @@
+/*
+ * 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.solr.client.solrj.impl;
+
+import org.apache.solr.client.solrj.SolrQuery;
+import org.apache.solr.client.solrj.request.CollectionAdminRequest;
+import org.apache.solr.client.solrj.response.CollectionAdminResponse;
+import org.apache.solr.client.solrj.response.QueryResponse;
+import org.apache.solr.cloud.SolrCloudTestCase;
+import org.apache.solr.common.SolrInputDocument;
+import org.junit.After;
+import org.junit.Before;
+import org.junit.Test;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.lang.invoke.MethodHandles;
+import java.util.ArrayList;
+import java.util.List;
+
+public class HttpClusterStateProviderTest extends SolrCloudTestCase {
+ private static final Logger log = LoggerFactory.getLogger(MethodHandles.lookup().lookupClass());
+ private static final int NODE_COUNT = 1;
+
+ private CloudSolrClient cloudSolrClient;
+ private CountingHttpClusterStateProvider httpClusterStateProvider;
+
+ @Before
+ public void setupCluster() throws Exception {
+ configureCluster(NODE_COUNT)
+ .addConfig("conf", getFile("solrj").toPath().resolve("solr").resolve("configsets").resolve("streaming").resolve("conf"))
+ .configure();
+
+ final List<String> solrUrls = new ArrayList<>();
+ solrUrls.add(cluster.getJettySolrRunner(0).getBaseUrl().toString());
+
+ httpClusterStateProvider = new CountingHttpClusterStateProvider(solrUrls, null);
+ cloudSolrClient = new CloudSolrClient(new CloudSolrClient.Builder(httpClusterStateProvider));
+ CollectionAdminRequest.Create create = CollectionAdminRequest.Create.createCollection("x", 1, 1);
+ CollectionAdminResponse adminResponse = create.process(cloudSolrClient);
+ assertTrue(adminResponse.isSuccess());
+ cloudSolrClient.setDefaultCollection("x");
+ }
+
+ @After
+ public void tearDown() throws Exception {
+ cloudSolrClient.close();
+ httpClusterStateProvider.close();
+ super.tearDown();
+ }
+
+ @Test
+ public void test() throws Exception {
+ // the constructor of HttpClusterStateProvider fetches live nodes
+ // and creating a collection fetches the cluster state
+ // so we can expect exactly 2 http calls to have been made already at this point
+ assertEquals(2, httpClusterStateProvider.getRequestCount());
+
+ QueryResponse queryResponse = cloudSolrClient.query(new SolrQuery("*:*"));
+ assertEquals(0, queryResponse.getResults().getNumFound());
+ // we can expect 1 extra call to fetch and cache the collection state
+ assertEquals(3, httpClusterStateProvider.getRequestCount());
+ queryResponse = cloudSolrClient.query(new SolrQuery("*:*"));
+ assertEquals(0, queryResponse.getResults().getNumFound());
+ // the collection state should already be in the cache so we do not expect another call
Review comment:
We expect this test to be running quickly, but if it slows for some reason (I'm thinking underpowered CI hosts) then it will fail with hard to debug state. Can we somehow set the cache time to very long for the context of this test?
##########
File path: solr/solrj/src/java/org/apache/solr/client/solrj/impl/BaseCloudSolrClient.java
##########
@@ -1226,7 +1227,7 @@ protected DocCollection getDocCollection(String collection, Integer expectedVers
//no such collection exists
return null;
}
- if (!ref.isLazilyLoaded()) {
+ if (getClusterStateProvider() instanceof ZkClientClusterStateProvider && !ref.isLazilyLoaded()) {
Review comment:
I don't understand why this is important, can you add to the comment?
##########
File path: solr/solrj/src/test/org/apache/solr/client/solrj/impl/CountingHttpClusterStateProvider.java
##########
@@ -0,0 +1,72 @@
+/*
+ * 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.solr.client.solrj.impl;
+
+import org.apache.http.client.HttpClient;
+import org.apache.solr.client.solrj.ResponseParser;
+import org.apache.solr.client.solrj.SolrClient;
+import org.apache.solr.client.solrj.SolrRequest;
+import org.apache.solr.client.solrj.SolrServerException;
+import org.apache.solr.common.util.NamedList;
+
+import java.io.IOException;
+import java.util.List;
+import java.util.concurrent.atomic.AtomicInteger;
+
+@SuppressWarnings({"unchecked"})
+public class CountingHttpClusterStateProvider extends BaseHttpClusterStateProvider {
+
+ private final HttpClient httpClient;
+ private final boolean clientIsInternal;
+
+ private final AtomicInteger counter = new AtomicInteger(0);
+
+ public CountingHttpClusterStateProvider(List<String> solrUrls, HttpClient httpClient) throws Exception {
+ this.httpClient = httpClient == null ? HttpClientUtil.createClient(null) : httpClient;
+ this.clientIsInternal = httpClient == null;
+ init(solrUrls);
+ }
+
+ @Override
+ protected SolrClient getSolrClient(String baseUrl) {
+ return new AssertingHttpSolrClient(new HttpSolrClient.Builder().withBaseSolrUrl(baseUrl).withHttpClient(httpClient));
+ }
+
+ @Override
+ public void close() throws IOException {
Review comment:
not thrown
----------------------------------------------------------------
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:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org