11
votes

We use apache ignite v2.2 as hibernate 2nd level cache in grails application. We have 4 nodes cluster with 10G RAM each. The first node starts ok. But subsequent hangs. Sometimes 2nd sometimes 3rd or 4th. Also successful startups happen but very rare. App hangs always in the same place:

"host-startStop-1" #45 daemon prio=5 os_prio=0 tid=0x00007f7cac004800 nid=0x3d44 waiting on condition [0x00007f7cfdd81000]
   java.lang.Thread.State: TIMED_WAITING (parking)
        at sun.misc.Unsafe.park(Native Method)
        at java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:338)
        at org.apache.ignite.internal.util.future.GridFutureAdapter.get0(GridFutureAdapter.java:216)
        at org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:158)
        at org.apache.ignite.internal.util.future.GridFutureAdapter.get(GridFutureAdapter.java:150)
        at org.apache.ignite.internal.processors.cache.GridCachePartitionExchangeManager.onKernalStart(GridCachePartitionExchangeManager.java:551)
        at org.apache.ignite.internal.processors.cache.GridCacheProcessor.onKernalStart(GridCacheProcessor.java:843)
        at org.apache.ignite.internal.IgniteKernal.start(IgniteKernal.java:1040)
        at org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start0(IgnitionEx.java:1896)
        at org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance.start(IgnitionEx.java:1648)
        - locked <0x00000007890a1198> (a org.apache.ignite.internal.IgnitionEx$IgniteNamedInstance)
        at org.apache.ignite.internal.IgnitionEx.start0(IgnitionEx.java:1076)
        at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:596)
        at org.apache.ignite.internal.IgnitionEx.start(IgnitionEx.java:520)
        at org.apache.ignite.Ignition.start(Ignition.java:322)

All other nodes are locked during this process. Configuration:

IgniteConfiguration configuration = new IgniteConfiguration()
        List<CacheConfiguration> cacheConfigurations = []
        for (String name : caches) {
            CacheConfiguration cacheConfiguration = new CacheConfiguration<>()
            cacheConfiguration.setCacheMode(CacheMode.REPLICATED)
            cacheConfiguration.setAtomicityMode(CacheAtomicityMode.TRANSACTIONAL)
            cacheConfiguration.setWriteSynchronizationMode(CacheWriteSynchronizationMode.FULL_ASYNC)
            cacheConfiguration.setName(name)
            cacheConfiguration.onheapCacheEnabled =  true
            cacheConfiguration.evictionPolicy = new LruEvictionPolicy()
            cacheConfiguration.memoryPolicyName = MEMORY_POLICY
            cacheConfigurations.add(cacheConfiguration)
        }
        for (String name : ['org.hibernate.cache.spi.UpdateTimestampsCache',
                            'org.hibernate.cache.internal.StandardQueryCache']) {
            CacheConfiguration cacheConfiguration = new CacheConfiguration<>()
            cacheConfiguration.setCacheMode(CacheMode.REPLICATED)
            cacheConfiguration.setAtomicityMode(CacheAtomicityMode.ATOMIC)
            cacheConfiguration.setWriteSynchronizationMode(CacheWriteSynchronizationMode.FULL_ASYNC)
            cacheConfiguration.setName(name)
            cacheConfiguration.onheapCacheEnabled =  true
            cacheConfiguration.evictionPolicy = new LruEvictionPolicy()
            cacheConfiguration.memoryPolicyName = MEMORY_POLICY
            cacheConfigurations.add(cacheConfiguration)
        }
        configuration.setCacheConfiguration(cacheConfigurations.toArray(new CacheConfiguration[cacheConfigurations.size()]))
        configuration.peerClassLoadingEnabled = true
        configuration.igniteInstanceName = Constants.IGNITE_GRID
        configuration.gridLogger = new Slf4jLogger()
        MemoryConfiguration memoryConfiguration = new MemoryConfiguration()
        memoryConfiguration.defaultMemoryPolicySize = 1 * 1024 * 1024 * 1024l
        MemoryPolicyConfiguration l2CachePolicy = new MemoryPolicyConfiguration()
        l2CachePolicy.name = MEMORY_POLICY
        l2CachePolicy.setMaxSize(4 * 1024 * 1024 * 1024l)
        l2CachePolicy.pageEvictionMode = DataPageEvictionMode.RANDOM_LRU
        memoryConfiguration.setMemoryPolicies(l2CachePolicy)
        configuration.memoryConfiguration = memoryConfiguration
        int[] eventTypes = new int[1]
        eventTypes[0] = EventType.EVT_NODE_FAILED
        configuration.includeEventTypes = eventTypes
        Map<IgnitePredicate<? extends Event>, int[]> listeners = new HashedMap()
        listeners.put(new NodeFailedEventListener(), eventTypes)
        configuration.localEventListeners = listeners
        TcpCommunicationSpi commSpi = new TcpCommunicationSpi()
        commSpi.slowClientQueueLimit = 1000
        commSpi.messageQueueLimit = 5000
        configuration.communicationSpi = commSpi
        TcpDiscoverySpi discoverySpi = new TcpDiscoverySpi()
        configuration.discoverySpi = discoverySpi
        if (grailsApplication.config.grails?.plugin?.awssdk?.accessKey && Env.igniteS3Bucket) {
            TcpDiscoveryS3IpFinder awsIpFinder = new TcpDiscoveryS3IpFinder()
            awsIpFinder.setBucketName(Env.igniteS3Bucket)
            AWSCredentials awsCredentials = new BasicAWSCredentials(grailsApplication.config.grails.plugin.awssdk.accessKey,
                    grailsApplication.config.grails.plugin.awssdk.secretKey)
            awsIpFinder.setAwsCredentials(awsCredentials)
            discoverySpi.ipFinder = awsIpFinder
        } else {
            TcpDiscoveryVmIpFinder ipFinder = new TcpDiscoveryVmIpFinder()
            ipFinder.setAddresses(["127.0.0.1:47500"])
            discoverySpi.ipFinder = ipFinder
        }
        configuration.classLoader = grailsApplication.classLoader
        ignite = Ignition.start(configuration)

EDIT

Full thread dump of failed node

Full thread dump of succeed node

2
It's impossible to understand anything from thread dump of one thread. Please share full logs and thread dumps from all nodesEvgenii Zhuravlev
I added full thread dumps of failed and succeeded nodesDmitry S

2 Answers

0
votes

If you want to run more than 1 node on one physical machine, I would recommend configuring MemoryConfiguration(because by default, in version 2.2 Ignite will require 80% of physical RAM for one node) or update to version 2.3(default value was reduced to 20%)

0
votes

Can you please try with caches that don't contain period '.' in their names? This is known to cause delayed problems.