magibney commented on code in PR #1899: URL: https://github.com/apache/solr/pull/1899#discussion_r2161511800
########## solr/core/src/java/org/apache/solr/search/OrdMapRegenerator.java: ########## @@ -0,0 +1,255 @@ +/* + * 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.search; + +import static org.apache.solr.common.params.CommonParams.NAME; + +import java.io.IOException; +import java.util.Collections; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.concurrent.TimeUnit; +import java.util.function.Function; +import java.util.function.Supplier; +import org.apache.lucene.index.DirectoryReader; +import org.apache.lucene.index.IndexReader; +import org.apache.lucene.index.LeafReaderContext; +import org.apache.lucene.index.OrdinalMap; +import org.apache.lucene.index.SortedDocValues; +import org.apache.lucene.index.SortedSetDocValues; +import org.apache.lucene.search.DocIdSetIterator; +import org.apache.lucene.util.RamUsageEstimator; +import org.apache.lucene.util.packed.PackedInts; +import org.apache.solr.core.SolrConfig; +import org.apache.solr.index.SlowCompositeReaderWrapper; +import org.apache.solr.search.SolrCache.MetaEntry; +import org.apache.solr.util.IOFunction; + +/** Cache regenerator that builds OrdinalMap instances against the new searcher. */ +public class OrdMapRegenerator<M extends MetaEntry<OrdinalMap, M>> + extends MetaCacheRegenerator<String, OrdinalMap, M> { + + private static final int DEFAULT_REGEN_KEEPALIVE_MINUTES = 2; + private static final long DEFAULT_REGEN_KEEPALIVE_NANOS = + TimeUnit.MINUTES.toNanos(DEFAULT_REGEN_KEEPALIVE_MINUTES); + private static final OrdMapRegenerator<?> DEFAULT_INSTANCE = + new OrdMapRegenerator<>(DEFAULT_REGEN_KEEPALIVE_NANOS); + + private final long regenKeepAliveNanos; + + public OrdMapRegenerator() { + this(DEFAULT_REGEN_KEEPALIVE_NANOS); + // default ctor in case someone specifies this class via standard `"regen"=[className]` syntax + } + + private OrdMapRegenerator(long regenKeepAliveNanos) { + super(getWrapFunction()); + this.regenKeepAliveNanos = regenKeepAliveNanos; + } + + @SuppressWarnings({"unchecked", "UnnecessaryLambda"}) + private static <M> Function<OrdinalMap, M> getWrapFunction() { + return (v) -> (M) new OrdinalMapValue(v, 0); + } + + public static class OrdinalMapValue implements MetaEntry<OrdinalMap, OrdinalMapValue> { + private static final long BASE_RAM_BYTES_USED = + RamUsageEstimator.shallowSizeOfInstance(OrdinalMapValue.class); + private final OrdinalMap ordinalMap; + private long accessTimestampNanos; + + private OrdinalMapValue(OrdinalMap ordinalMap, long accessTimestampNanos) { + this.ordinalMap = ordinalMap; + this.accessTimestampNanos = accessTimestampNanos; + } + + @Override + public OrdinalMap get() { + accessTimestampNanos = System.nanoTime(); + return ordinalMap; + } + + @Override + public long ramBytesUsed() { + return BASE_RAM_BYTES_USED + ordinalMap.ramBytesUsed(); + } + + @Override + public OrdinalMapValue metaClone(OrdinalMap val) { + return new OrdinalMapValue(val, accessTimestampNanos); + } + } + + /** + * Configures and supplies a default `ordMapCache` {@link CacheConfig} of unlimited maximum size + * and modest initial size, with no autowarming. + */ + public static CacheConfig getDefaultCacheConfig() { + // for back-compat, default to an effectively unlimited-sized cache with no regeneration + Map<String, String> args = new HashMap<>(); + args.put(NAME, "ordMapCache"); + args.put("size", Integer.toString(Integer.MAX_VALUE)); // effectively unlimited + args.put("initialSize", "10"); + return new CacheConfig(CaffeineCache.class, args, null); + } + + /** + * If no regenerator is explicitly configured for the specified {@link CacheConfig}, and if + * autowarming is on, this method configures a regenerator. `regenKeepAlive` is respected if + * explicitly set, implicitly defaulting to 2x the autocommit interval defined by the specified + * {@link SolrConfig} (or default of {@value #DEFAULT_REGEN_KEEPALIVE_MINUTES} minutes if no + * autocommit interval can be determined). + */ + public static void configureRegenerator(SolrConfig solrConfig, CacheConfig config) { + if (config.getRegenerator() != null + || !new SolrCacheBase.AutoWarmCountRef( + (String) config.toMap(new HashMap<>()).get("autowarmCount")) + .isAutoWarmingOn()) { + // If a regenerator is already explicitly configured, we don't want to replace it. + // Also, if autowarming is not on, we don't configure a regenerator. This is important + // because the regenerator is also used to wrap/unwrap the cache for the purpose of + // tracking metadata, etc. If there's no autowarm, the extra overhead is useless. + return; + } + String keepAliveConfig = (String) config.toMap(Collections.emptyMap()).get("regenKeepAlive"); + final long regenKeepAliveNanos; + if (keepAliveConfig == null || keepAliveConfig.isEmpty()) { + long osiNanos; + if (solrConfig == null || (osiNanos = getOpenSearcherIntervalNanos(solrConfig)) == -1) { + regenKeepAliveNanos = DEFAULT_REGEN_KEEPALIVE_NANOS; + } else { + regenKeepAliveNanos = osiNanos << 1; + } + } else { + int lastIdx = keepAliveConfig.length() - 1; + String sub = keepAliveConfig.substring(0, lastIdx); + switch (keepAliveConfig.charAt(lastIdx)) { + case 's': + regenKeepAliveNanos = TimeUnit.SECONDS.toNanos(Long.parseLong(sub)); + break; + case 'm': + regenKeepAliveNanos = TimeUnit.MINUTES.toNanos(Long.parseLong(sub)); + break; + case 'h': + regenKeepAliveNanos = TimeUnit.HOURS.toNanos(Long.parseLong(sub)); + break; + case 'd': + regenKeepAliveNanos = TimeUnit.DAYS.toNanos(Long.parseLong(sub)); + break; + case '%': + int keepAlivePct = Integer.parseInt(sub); + if (keepAlivePct < 0) { + throw new IllegalArgumentException( + "regenKeepAlive % must be positive; found " + keepAlivePct); + } + long osiNanos; + if (solrConfig == null || (osiNanos = getOpenSearcherIntervalNanos(solrConfig)) == -1) { + throw new IllegalArgumentException( + "regenKeepAlive % must only be configured in conjunction with autoCommit time"); + } else { + regenKeepAliveNanos = (osiNanos * keepAlivePct) / 100; + } + break; + default: + regenKeepAliveNanos = TimeUnit.MILLISECONDS.toNanos(Long.parseLong(keepAliveConfig)); + break; + } + } + if (regenKeepAliveNanos == DEFAULT_REGEN_KEEPALIVE_NANOS) { + config.setRegenerator(DEFAULT_INSTANCE); + return; + } + config.setRegenerator(new OrdMapRegenerator<>(regenKeepAliveNanos)); + } + + private static long getOpenSearcherIntervalNanos(SolrConfig solrConfig) { + SolrConfig.UpdateHandlerInfo uinfo = solrConfig.getUpdateHandlerInfo(); + if (uinfo == null) { + return -1; + } else if (uinfo.autoSoftCommmitMaxTime != -1) { + if (uinfo.openSearcher && uinfo.autoCommmitMaxTime != -1) { + return TimeUnit.MILLISECONDS.toNanos( + Math.min(uinfo.autoCommmitMaxTime, uinfo.autoSoftCommmitMaxTime)); + } else { + return TimeUnit.MILLISECONDS.toNanos(uinfo.autoSoftCommmitMaxTime); + } + } else if (uinfo.openSearcher && uinfo.autoCommmitMaxTime != -1) { + return TimeUnit.MILLISECONDS.toNanos(uinfo.autoCommmitMaxTime); + } else { + return -1; + } + } + + @Override + public <K, V> boolean regenerateItem( + SolrIndexSearcher newSearcher, + SolrCache<K, V> newCache, + SolrCache<K, V> oldCache, + K oldKey, + V oldVal) + throws IOException { + DirectoryReader in = newSearcher.getIndexReader(); + IndexReader.CacheHelper cacheHelper = in.getReaderCacheHelper(); + if (cacheHelper == null) { + return false; + } + + final List<LeafReaderContext> leaves = in.leaves(); + final int size = leaves.size(); + + if (size < 2) { + // we don't need OrdinalMaps for these trivial cases + return false; + } + + OrdinalMapValue ordinalMapValue = (OrdinalMapValue) oldVal; + final long extantTimestamp = ordinalMapValue.accessTimestampNanos; + if (System.nanoTime() - extantTimestamp > regenKeepAliveNanos) { + // it has been long enough since this was last accessed that we don't want to carry it forward + return true; Review Comment: Shouldn't this return `false` instead? -- 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: issues-unsubscr...@solr.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org For additional commands, e-mail: issues-h...@solr.apache.org
