I'm programming an application with the latest version of Spring Boot. I recently became problems with growing heap, that can not be garbage collected. The analysis of the heap with Eclipse MAT showed that, within one hour of running the application, the heap grew to 630MB and with Hibernate's SessionFactoryImpl using more than 75% of the whole heap.
Is was looking for possible sources around the Query Plan Cache, but the only thing I found was this, but that did not play out. The properties were set like this:
spring.jpa.properties.hibernate.query.plan_cache_max_soft_references=1024
spring.jpa.properties.hibernate.query.plan_cache_max_strong_references=64
The database queries are all generated by the Spring's Query magic, using repository interfaces like in this documentation. There are about 20 different queries generated with this technique. No other native SQL or HQL are used.
Sample:
#Transactional
public interface TrendingTopicRepository extends JpaRepository<TrendingTopic, Integer> {
List<TrendingTopic> findByNameAndSource(String name, String source);
List<TrendingTopic> findByDateBetween(Date dateStart, Date dateEnd);
Long countByDateBetweenAndName(Date dateStart, Date dateEnd, String name);
}
or
List<SomeObject> findByNameAndUrlIn(String name, Collection<String> urls);
as example for IN usage.
Question is: Why does the query plan cache keep growing (it does not stop, it ends in a full heap) and how to prevent this? Did anyone encounter a similar problem?
Versions:
Spring Boot 1.2.5
Hibernate 4.3.10
I've hit this issue as well. It basically boils down to having variable number of values in your IN clause and Hibernate trying to cache those query plans.
There are two great blog posts on this topic.
The first:
Using Hibernate 4.2 and MySQL in a project with an in-clause query
such as: select t from Thing t where t.id in (?)
Hibernate caches these parsed HQL queries. Specifically the Hibernate
SessionFactoryImpl has QueryPlanCache with queryPlanCache and
parameterMetadataCache. But this proved to be a problem when the
number of parameters for the in-clause is large and varies.
These caches grow for every distinct query. So this query with 6000
parameters is not the same as 6001.
The in-clause query is expanded to the number of parameters in the
collection. Metadata is included in the query plan for each parameter
in the query, including a generated name like x10_, x11_ , etc.
Imagine 4000 different variations in the number of in-clause parameter
counts, each of these with an average of 4000 parameters. The query
metadata for each parameter quickly adds up in memory, filling up the
heap, since it can't be garbage collected.
This continues until all different variations in the query parameter
count is cached or the JVM runs out of heap memory and starts throwing
java.lang.OutOfMemoryError: Java heap space.
Avoiding in-clauses is an option, as well as using a fixed collection
size for the parameter (or at least a smaller size).
For configuring the query plan cache max size, see the property
hibernate.query.plan_cache_max_size, defaulting to 2048 (easily too
large for queries with many parameters).
And second (also referenced from the first):
Hibernate internally uses a cache that maps HQL statements (as
strings) to query plans. The cache consists of a bounded map limited
by default to 2048 elements (configurable). All HQL queries are loaded
through this cache. In case of a miss, the entry is automatically
added to the cache. This makes it very susceptible to thrashing - a
scenario in which we constantly put new entries into the cache without
ever reusing them and thus preventing the cache from bringing any
performance gains (it even adds some cache management overhead). To
make things worse, it is hard to detect this situation by chance - you
have to explicitly profile the cache in order to notice that you have
a problem there. I will say a few words on how this could be done
later on.
So the cache thrashing results from new queries being generated at
high rates. This can be caused by a multitude of issues. The two most
common that I have seen are - bugs in hibernate which cause parameters
to be rendered in the JPQL statement instead of being passed as
parameters and the use of an "in" - clause.
Due to some obscure bugs in hibernate, there are situations when
parameters are not handled correctly and are rendered into the JPQL
query (as an example check out HHH-6280). If you have a query that is
affected by such defects and it is executed at high rates, it will
thrash your query plan cache because each JPQL query generated is
almost unique (containing IDs of your entities for example).
The second issue lays in the way that hibernate processes queries with
an "in" clause (e.g. give me all person entities whose company id
field is one of 1, 2, 10, 18). For each distinct number of parameters
in the "in"-clause, hibernate will produce a different query - e.g.
select x from Person x where x.company.id in (:id0_) for 1 parameter,
select x from Person x where x.company.id in (:id0_, :id1_) for 2
parameters and so on. All these queries are considered different, as
far as the query plan cache is concerned, resulting again in cache
thrashing. You could probably work around this issue by writing a
utility class to produce only certain number of parameters - e.g. 1,
10, 100, 200, 500, 1000. If you, for example, pass 22 parameters, it
will return a list of 100 elements with the 22 parameters included in
it and the remaining 78 parameters set to an impossible value (e.g. -1
for IDs used for foreign keys). I agree that this is an ugly hack but
could get the job done. As a result you will only have at most 6
unique queries in your cache and thus reduce thrashing.
So how do you find out that you have the issue? You could write some
additional code and expose metrics with the number of entries in the
cache e.g. over JMX, tune logging and analyze the logs, etc. If you do
not want to (or can not) modify the application, you could just dump
the heap and run this OQL query against it (e.g. using mat): SELECT l.query.toString() FROM INSTANCEOF org.hibernate.engine.query.spi.QueryPlanCache$HQLQueryPlanKey l. It
will output all queries currently located in any query plan cache on
your heap. It should be pretty easy to spot whether you are affected
by any of the aforementioned problems.
As far as the performance impact goes, it is hard to say as it depends
on too many factors. I have seen a very trivial query causing 10-20 ms
of overhead spent in creating a new HQL query plan. In general, if
there is a cache somewhere, there must be a good reason for that - a
miss is probably expensive so your should try to avoid misses as much
as possible. Last but not least, your database will have to handle
large amounts of unique SQL statements too - causing it to parse them
and maybe create different execution plans for every one of them.
I have same problems with many(>10000) parameters in IN-queries. The number of my parameters is always different and I can not predict this, my QueryCachePlan growing too fast.
For database systems supporting execution plan caching, there's a better chance of hitting the cache if the number of possible IN clause parameters lowers.
Fortunately Hibernate of version 5.2.18 and higher has a solution with padding of parameters in IN-clause.
Hibernate can expand the bind parameters to power-of-two: 4, 8, 16, 32, 64.
This way, an IN clause with 5, 6, or 7 bind parameters will use the 8 IN clause, therefore reusing its execution plan.
If you want to activate this feature, you need to set this property to true hibernate.query.in_clause_parameter_padding=true.
For more information see this article, atlassian.
I had the exact same problem using Spring Boot 1.5.7 with Spring Data (Hibernate) and the following config solved the problem (memory leak):
spring:
jpa:
properties:
hibernate:
query:
plan_cache_max_size: 64
plan_parameter_metadata_max_size: 32
Starting with Hibernate 5.2.12, you can specify a hibernate configuration property to change how literals are to be bound to the underlying JDBC prepared statements by using the following:
hibernate.criteria.literal_handling_mode=BIND
From the Java documentation, this configuration property has 3 settings
AUTO (default)
BIND - Increases the likelihood of jdbc statement caching using bind parameters.
INLINE - Inlines the values rather than using parameters (be careful of SQL injection).
I had a similar issue, the issue is because you are creating the query and not using the PreparedStatement. So what happens here is for each query with different parameters it creates an execution plan and caches it.
If you use a prepared statement then you should see a major improvement in the memory being used.
TL;DR: Try to replace the IN() queries with ANY() or eliminate them
Explanation:
If a query contains IN(...) then a plan is created for each amount of values inside IN(...), since the query is different each time.
So if you have IN('a','b','c') and IN ('a','b','c','d','e') - those are two different query strings/plans to cache. This answer tells more about it.
In case of ANY(...) a single (array) parameter can be passed, so the query string will remain the same and the prepared statement plan will be cached once (example given below).
Cause:
This line might cause the issue:
List<SomeObject> findByNameAndUrlIn(String name, Collection<String> urls);
as under the hood it generates different IN() queries for every amount of values in "urls" collection.
Warning:
You may have IN() query without writing it and even without knowing about it.
ORM's such as Hibernate may generate them in the background - sometimes in unexpected places and sometimes in a non-optimal ways.
So consider enabling query logs to see the actual queries you have.
Fix:
Here is a (pseudo)code that may fix issue:
query = "SELECT * FROM trending_topic t WHERE t.name=? AND t.url=?"
PreparedStatement preparedStatement = connection.prepareStatement(queryTemplate);
currentPreparedStatement.setString(1, name); // safely replace first query parameter with name
currentPreparedStatement.setArray(2, connection.createArrayOf("text", urls.toArray())); // replace 2nd parameter with array of texts, like "=ANY(ARRAY['aaa','bbb'])"
But:
Don't take any solution as a ready-to-use answer. Make sure to test the final performance on actual/big data before going to production - no matter which answer you choose.
Why? Because IN and ANY both have pros and cons, and they can bring serious performance issues if used improperly (see examples in references below). Also make sure to use parameter binding to avoid security issues as well.
References:
100x faster Postgres performance by changing 1 line - performance of Any(ARRAY[]) vs ANY(VALUES())
Index not used with =any() but used with in - different performance of IN and ANY
Understanding SQL Server query plan cache
Hope this helps. Make sure to leave a feedback whether it worked or not - in order to help people like you. Thanks!
I had a big issue with this queryPlanCache, so I did a Hibernate cache monitor to see the queries in the queryPlanCache.
I am using in QA environment as a Spring task each 5 minutes.
I found which IN queries I had to change to solve my cache problem.
A detail is: I am using Hibernate 4.2.18 and I don't know if will be useful with other versions.
import java.lang.reflect.Field;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
import java.util.Set;
import javax.persistence.EntityManager;
import javax.persistence.PersistenceContext;
import org.hibernate.ejb.HibernateEntityManagerFactory;
import org.hibernate.internal.SessionFactoryImpl;
import org.hibernate.internal.util.collections.BoundedConcurrentHashMap;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import com.dao.GenericDAO;
public class CacheMonitor {
private final Logger logger = LoggerFactory.getLogger(getClass());
#PersistenceContext(unitName = "MyPU")
private void setEntityManager(EntityManager entityManager) {
HibernateEntityManagerFactory hemf = (HibernateEntityManagerFactory) entityManager.getEntityManagerFactory();
sessionFactory = (SessionFactoryImpl) hemf.getSessionFactory();
fillQueryMaps();
}
private SessionFactoryImpl sessionFactory;
private BoundedConcurrentHashMap queryPlanCache;
private BoundedConcurrentHashMap parameterMetadataCache;
/*
* I tried to use a MAP and use compare compareToIgnoreCase.
* But remember this is causing memory leak. Doing this
* you will explode the memory faster that it already was.
*/
public void log() {
if (!logger.isDebugEnabled()) {
return;
}
if (queryPlanCache != null) {
long cacheSize = queryPlanCache.size();
logger.debug(String.format("QueryPlanCache size is :%s ", Long.toString(cacheSize)));
for (Object key : queryPlanCache.keySet()) {
int filterKeysSize = 0;
// QueryPlanCache.HQLQueryPlanKey (Inner Class)
Object queryValue = getValueByField(key, "query", false);
if (queryValue == null) {
// NativeSQLQuerySpecification
queryValue = getValueByField(key, "queryString");
filterKeysSize = ((Set) getValueByField(key, "querySpaces")).size();
if (queryValue != null) {
writeLog(queryValue, filterKeysSize, false);
}
} else {
filterKeysSize = ((Set) getValueByField(key, "filterKeys")).size();
writeLog(queryValue, filterKeysSize, true);
}
}
}
if (parameterMetadataCache != null) {
long cacheSize = parameterMetadataCache.size();
logger.debug(String.format("ParameterMetadataCache size is :%s ", Long.toString(cacheSize)));
for (Object key : parameterMetadataCache.keySet()) {
logger.debug("Query:{}", key);
}
}
}
private void writeLog(Object query, Integer size, boolean b) {
if (query == null || query.toString().trim().isEmpty()) {
return;
}
StringBuilder builder = new StringBuilder();
builder.append(b == true ? "JPQL " : "NATIVE ");
builder.append("filterKeysSize").append(":").append(size);
builder.append("\n").append(query).append("\n");
logger.debug(builder.toString());
}
private void fillQueryMaps() {
Field queryPlanCacheSessionField = null;
Field queryPlanCacheField = null;
Field parameterMetadataCacheField = null;
try {
queryPlanCacheSessionField = searchField(sessionFactory.getClass(), "queryPlanCache");
queryPlanCacheSessionField.setAccessible(true);
queryPlanCacheField = searchField(queryPlanCacheSessionField.get(sessionFactory).getClass(), "queryPlanCache");
queryPlanCacheField.setAccessible(true);
parameterMetadataCacheField = searchField(queryPlanCacheSessionField.get(sessionFactory).getClass(), "parameterMetadataCache");
parameterMetadataCacheField.setAccessible(true);
queryPlanCache = (BoundedConcurrentHashMap) queryPlanCacheField.get(queryPlanCacheSessionField.get(sessionFactory));
parameterMetadataCache = (BoundedConcurrentHashMap) parameterMetadataCacheField.get(queryPlanCacheSessionField.get(sessionFactory));
} catch (Exception e) {
logger.error("Failed fillQueryMaps", e);
} finally {
queryPlanCacheSessionField.setAccessible(false);
queryPlanCacheField.setAccessible(false);
parameterMetadataCacheField.setAccessible(false);
}
}
private <T> T getValueByField(Object toBeSearched, String fieldName) {
return getValueByField(toBeSearched, fieldName, true);
}
#SuppressWarnings("unchecked")
private <T> T getValueByField(Object toBeSearched, String fieldName, boolean logErro) {
Boolean accessible = null;
Field f = null;
try {
f = searchField(toBeSearched.getClass(), fieldName, logErro);
accessible = f.isAccessible();
f.setAccessible(true);
return (T) f.get(toBeSearched);
} catch (Exception e) {
if (logErro) {
logger.error("Field: {} error trying to get for: {}", fieldName, toBeSearched.getClass().getName());
}
return null;
} finally {
if (accessible != null) {
f.setAccessible(accessible);
}
}
}
private Field searchField(Class<?> type, String fieldName) {
return searchField(type, fieldName, true);
}
private Field searchField(Class<?> type, String fieldName, boolean log) {
List<Field> fields = new ArrayList<Field>();
for (Class<?> c = type; c != null; c = c.getSuperclass()) {
fields.addAll(Arrays.asList(c.getDeclaredFields()));
for (Field f : c.getDeclaredFields()) {
if (fieldName.equals(f.getName())) {
return f;
}
}
}
if (log) {
logger.warn("Field: {} not found for type: {}", fieldName, type.getName());
}
return null;
}
}
We also had a QueryPlanCache with growing heap usage. We had IN-queries which we rewrote, and additionally we have queries which use custom types. Turned out that the Hibernate class CustomType didn't properly implement equals and hashCode thereby creating a new key for every query instance. This is now solved in Hibernate 5.3.
See https://hibernate.atlassian.net/browse/HHH-12463.
You still need to properly implement equals/hashCode in your userTypes to make it work properly.
We had faced this issue with query plan cache growing too fast and old gen heap was also growing along with it as gc was unable to collect it.The culprit was JPA query taking some more than 200000 ids in the IN clause. To optimise the query we used joins instead of fetching ids from one table and passing those in other table select query..
I started using compiled queries to increase the performance of some commonly executed linq to entities queries. In one scenario I only boiled the query down to it's most basic form and pre-compiled that, then I tack on additional where clauses based on user input.
I seem to be losing the performance benefit of compiled queries in this particular case. Can someone explain why?
Here's an example of what I'm doing...
IEnumerable<Task> tasks = compiledQuery.Invoke(context, userId);
if(status != null)
{
tasks = tasks.Where(x=x.Status == status);
}
if(category != null)
{
tasks = tasks.Where(x=x.Category == category);
}
return tasks;
I think it's important to understand how Compiled Queries in EF work.
When you execute a query Entity Framework will map your expression tree with the help of your mapping file (EDMX or with code first your model definitions) to a SQL query. This can be a complex and performance intensive task.
Precompiling stores the results of these mapping phase so the next time you hit the query it has the SQL already available and it only has to set the current parameters.
The problem is that a precompiled query will lose it's performance benefit as soon as you modifie the query. Let's say you have the following:
IQueryable query = GetCompiledQuery(); // => db.Tasks.Where(t => t.Id == myId);
var notModifiedResult = query.ToList(); // Fast
int ModifiedResult = query.Count(); // Slow
With the first query you will have all the benefits of precompiling because EF has the SQL already generated for you and can execute this immediatly.
The second query will lose the precompiling because it has to regenerate it's SQL.
If you would now execute a query on notModifiedResult this will be a Linq To Objects one because you have already executed your SQL to the database and fetched all the elements in memory.
You can however chain Compiled Queries (that is, use a compiled query in another compiled query).
But your code would require a series of compiled queries:
- The default
- One where status != null
- One where category != null
- One where both status and category != null
(Note: I haven't done any EF work for ages, and then it was just pottering. This is just an informed guess, really.)
This could be the culprit:
IEnumerable<Task> tasks = compiledQuery.Invoke(context, userId);
Any further querying will have to be done within the .NET process, not in SQL. All the possible results will have to be fetched from the database and filtered locally. Try this instead:
IQueryable<Task> tasks = compiledQuery.Invoke(context, userId);
(Assuming that's valid, of course.)
The compiled query can't be changed, only the parameters can be changed. What you are doing here is actually running the query, and THEN filtering the results.
.Invoke(context, userId); // returns all the results
.Where(....) // filters on that entire collection
You can see if there is a clever way to restate your query, so that the parameters can be included in all cases, but not have any effect. I haven't worked with compiled queries, sorry about that, but does this work (using -1 as the "ignore" value)?
// bunch of code to define the compiled query part, copied from [msdn][1]
(ctx, total) => from order in ctx.SalesOrderHeaders
where (total == -1 || order.TotalDue >= total)
select order);
In SQL, you do this by either using dynamic sql, or having a default value (or null) that you pass in which indicates that parameter should be ignored
select * from table t
where
(#age = 0 or t.age = #age) and
(#weight is null or t.weight = #weight)
I am writing an application that works with MS SQL database via LINQ to SQL. I need to perform filtering sometimes, and occasionally my filtering conditions are too complicated to be translated into SQL query. While I am trying to make them translatable, I want my application to at least work, though slow sometimes.
LINQ to SQL data model is hidden inside repositories, and I do not want to provide several GetAll method overloads for different cases and be aware of what overload to use on upper levels. So I want to test my expression inside repository to be translatable and, if no, perform in-memory query against the whole data set instead of throwing NotSupportedException on query instantiating.
This is what I have now:
IQueryable<TEntity> table = GetTable<TEntity>();
IQueryable<TEntity> result;
try
{
result = table.Where(searchExpression);
//this will test our expression
//consuming as little resources as possible (???)
result.FirstOrDefault();
}
catch (NotSupportedException)
{
//trying to perform in-memory search if query could not be constructed
result = table
.AsEnumerable()
.Where(searchExpression.Compile())
.AsQueryable();
}
return result;
searchExpression is Expression<Func<TEntity, bool>>
As you see, I am using FirstOrDefault to try to instantiate the query and throw the exception if it cannot be instantiated. However, it will perform useless database call when the expression is good. I could use Any, Count or other method, and it may well be a bit less expensive then FirstOrDefault, but still all methods that come to my mind make a costly trip to database, while all I need is to test my expression.
Is there any alternative way to say whether my expression is 'good' or 'bad', without actual database call?
UPDATE:
Or, more generally, is there a way to tell LINQ to make in-memory queries when it fails to construct SQL, so that this testing mechanism would not be needed at all?
Instead of
result.FirstOrDefault();
would it be sufficient to use
string sqlCommand = dataContext.GetCommand(result).CommandText;
?
If the expression does not generate valid Sql, this should throw a NotSupportedException, but it does not actually execute the sqlCommand.
I think this will solve your problem:
IQueryable<TEntity> table = GetTable<TEntity>();
IQueryable<TEntity> result;
try
{
return table.Where(searchExpression).ToList();
}
catch (NotSupportedException)
{
//trying to perform in-memory search if query could not be constructed
return table
.AsEnumerable()
.Where(searchExpression.Compile())
.ToList();
}
So the method returns is the expression is converted to valid SQL. Otherwise it catches the exception and runs the query in memory. This should work but it doesn't answer your question if it's possible to check if a specific searchExpression can be converted. I don't think such a thing exists.
i have a problem with some data i retrievied from db with linq.
When I try to access data I obtain the following exception:
System.ObjectDisposedException : The istance of ObjectContext was deleted and is not possible to use it again for action that need a connection.
This is the code:
using (ProvaDbEntities DBEntities =
new ProvaDbEntities(Utilities.ToEntitiesConnectionString()))
{
ObjectQuery<site> sites = DBEntities.site;
IEnumerable<site> q = from site in sites
select site;
{
ObjectQuery<auction> auctions = DBEntities.auction;
IEnumerable<auction> q1 = from auction in auctions
where auction.site == this.Name
select auction;
IEnumerable<IAuction> res = q1.Cast<IAuction>();
return res;
}
}
catch(Exception e)
{
throw new UnavailableDbException("[GetAuctions]" + e.Message);
}
Someone can help me???
Tanks
Fabio
Yes - you're returning a result which will be lazily evaluated - but you're disposing of the data context which would be used to fetch the results.
Options:
Load the results eagerly, e.g. by calling ToList on the result
Don't dispose of the context (I don't know what the situation is in the Entity Framework; you could get away with this in LINQ to SQL, but it may not be a good idea in EF)
Dispose of the context when you're finished with the data
In this case I'd suggest using the first option - it'll be safe and simple. As you're already filtering the results and you're casting to IEnumerable<IAuction> anyway, you're unlikely to get the normal downsides of materializing the query early. (If it were still IQueryable<T>, you'd be throwing away the ability to add extra bits to the query and them still be translated to SQL.)
If I were to select some rows based on certain criteria I can use ICriterion object in NHibernate.Criterion, such as this:
public List<T> GetByCriteria()
{
SimpleExpression newJobCriterion =
NHibernate.Criterion.Expression.Eq("LkpStatu", statusObject);
ICriteria criteria = Session.GetISession().CreateCriteria(typeof(T)).SetMaxResults(maxResults);
criteria.Add(newJobCriterion );
return criteria.List<T>();
}
Or I can use LINQ's where clause to filter what I want:
public List<T> GetByCriteria_LINQ()
{
ICriteria criteria = Session.GetISession().CreateCriteria(typeof(T)).SetMaxResults(maxResults);
return criteria.Where(item=>item.LkpStatu=statusObject).ToList();
}
I would prefer the second one, of course. Because
It gives me strong typing
I don't need to learn yet-another-syntax in the form of NHibernate
The issue is is there any performance advantage of the first one over the second one? From what I know, the first one will create SQL queries, so it will filter the data before pass into the memory. Is this kind of performance saving big enough to justify its use?
As usual it depends. First note that in your second snippet there is .List() missing right after return criteria And also note that you won't get the same results on both examples. The first one does where and then return top maxResults, the second one however first selects top maxResults and then does where.
If your expected result set is relatively small and you are likely to use some of the results in lazy loads then it's actually better to take the second approach. Because all entities loaded through a session will stay in its first level cache.
Usually however you don't do it this way and use the first approach.
Perhaps you wanted to use NHibernate.Linq (located in Contrib project ). Which does linq translation to Criteria for you.
I combine the two and made this:
var crit = _session.CreateCriteria(typeof (T)).SetMaxResults(100);
return (from x in _session.Linq<T>(crit) where x.field == <something> select x).ToList();