Apache Ignite Near Cache 总是错过

Question

当使用近缓存时，一切都会正常工作，直到第二个客户端（可能是 visor）在缓存操作正在进行时尝试连接或断开与集群的连接。

第二个客户端连接/断开连接后，原来的客户端总是会错过near cache，直到原来的客户端重新启动。几乎就像集群通知客户他们的问题并保持集群作为事实的来源一样。

我们已经能够通过运行我们的测试并连接/断开与 visor 的连接来重现这一点。在断开连接期间，我们可以在原始客户端 IgniteTxManager$NodeFailureTimeoutObject 的日志中看到提到的超时。

下面是 org.apache.ignite.internal.processors 被抑制的日志的 sn-p。

[2017-10-09 14:26:52.148] boot - 9955 DEBUG [http-nio-8081-exec-8] --- CacheHelper: 访问缓存 ng-security-service-ORG_SPEC_CACHE 的总时间 * |值 com.cache.model.PrefixCluster@78475a88: 0 毫秒 [2017-10-09 14:26:52.150] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDiscoveryManager：守护程序节点左侧拓扑：TcpDiscoveryNode [id=4cc6c321-d9cc-4149-a6ef -cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17 .0.1:0, /127.0.0.1:0, /10.70.255.8:0], discPort=0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1: a6ca5c8a, isClient=true] [2017-10-09 14:26:52.150] 启动 - 9955 调试 [http-nio-8081-exec-8] --- OrgSpecCacheImpl：OrgSpec 缓存统计信息：OrgSpec ObjId：IgniteCacheProxy [delegate=GridNearCacheAdapter []，opCtx=null , restartFut=null] HitCount: 120, MissCount: 50, AvgReadTime: 120，驱逐计数：0 [2017-10-09 14:26:52.150] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentPerVersionStore：处理节点离开事件：DiscoveryEvent [evtNode=TcpDiscoveryNode [id=4cc6c321-d9cc -4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17.0.1:0, /127.0.0.1:0, /10.70.255.8:0], discPort= 0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], topVer=58, nodeId8=2e573c60, msg=左节点：TcpDiscoveryNode [ id=4cc6c321-d9cc-4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17.0.1:0, /127.0.0.1:0, /10.70.255.8: 0], discPort=0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], type=NODE_LEFT, tstamp=1507577212142] [2017-10-09 14:26:52.163] 启动 - 9955 INFO [exchange-worker-#27%null%] --- 时间：开始交换 init [topVer=AffinityTopologyVersion [topVer=58, minorTopVer=0], crd =false，evt=11，节点=TcpDiscoveryNode [id=2e573c60-45f0-4429-a3fa-068489663148，addrs=[0:0:0:0:0:0:0:1%lo, 10.70.242.138, 127.0。 0.1], sockAddrs=[port-svc-inc-13.tw-test.net/10.70.242.138:0, /0:0:0:0:0:0:0:1%lo:0, /127.0. 0.1:0], discPort=0, order=56, intOrder=0, lastExchangeTime=1507576971754, loc=true, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], evtNode=TcpDiscoveryNode [id=2e573c60- 45f0-4429-a3fa-068489663148, addrs=[0:0:0:0:0:0:0:1%lo, 10.70.242.138, 127.0.0.1], sockAddrs=[port-svc-inc-13.tw -test.net/10.70.242.138:0, /0:0:0:0:0:0:0:1%lo:0, /127.0.0.1:0], discPort=0, order=56, intOrder= 0, lastExchangeTime=1507576971754, loc=true, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], customEvt=null] [2017-10-09 14:26:52.164] 启动 - 9955 INFO [exchange-worker-#27%null%] --- GridDhtPartitionsExchangeFuture：快照初始化已完成 [topVer=AffinityTopologyVersion [topVer=58，minorTopVer=0]，时间=0毫秒] [2017-10-09 14:26:52.164] 启动 - 9955 INFO [exchange-worker-#27%null%] --- GridDhtPartitionsExchangeFuture：快照初始化已完成 [topVer=AffinityTopologyVersion [topVer=58，minorTopVer=0]，时间=0毫秒] [2017-10-09 14:26:52.164] 启动 - 9955 INFO [exchange-worker-#27%null%] --- 时间：完成交换初始化 [topVer=AffinityTopologyVersion [topVer=58, minorTopVer=0], crd =假] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：本地部署的部署元数据：GridDeploymentMetadata [depMode=SHARED，别名=org。 apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1, clsName=org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1, userVer=null, sndNodeId=2e573c60-45f0-4429-a3fa-068489663148, clsLdrId=null，clsLdr=null，参与者=null，parentLdr=null，record=true，nodeFilter=null，seqNum=n/a] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：注册 [ldrRsrcs={org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f ={org.apache.ignite.internal.util.typedef.T2=org.apache.ignite.internal.util.typedef.T2, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap= org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap，org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionFullMap=org.apache.ignite.internal.processors。 cache.distributed.dht.preloader.GridDhtPartitionFullMap，java.util.Collections$UnmodifiableList=java.util.Collections$UnmodifiableList，org.apache.ignite.internal.visor.cache.VisorCacheMetricsCollectorTask=org.apache.ignite.internal.visor。 cache.VisorCacheMetricsCollectorTask, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMap =org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMap, org.apache.ignite.internal.processors.service.GridServiceProcessor$1=org.apache.ignite.internal.processors.service.GridServiceProcessor$1 , org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionMap=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionMap, org.apache.ignite.internal.processors .cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap}}, ldr=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f, rsrc=class org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：要注册的资源：{org.apache.ignite.internal.processors.task .GridTaskProcessor$TaskDiscoveryListener$1=org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1} [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：新资源：{org.apache.ignite.internal.processors.task. GridTaskProcessor$TaskDiscoveryListener$1=org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1} [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：删除资源 [clsLdrToIgnore=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f , rsrcs={org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1=org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1}] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：从 spi 检索自动加载的资源：DeploymentResourceAdapter [name=org.apache。 ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1, rsrcCls=class org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1, clsLdr=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：获取的部署类：GridDeployment [ts=1507576972855，depMode=SHARED，clsLdr= org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f, clsLdrId=6d9e6920f51-2e573c60-45f0-4429-a3fa-068489663148, userVer=0, loc=true, sampleClsName=org.apache.ignite.internal.processors.cache.distributed。 dht.preloader.GridDhtPartitionFullMap，pendingUndeploy=false，undeployed=false，usage=0] [2017-10-09 14:26:52.203] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridResourceProcessor：注入资源 [target=org.apache.ignite.internal.processors.task .GridTaskProcessor$TaskDiscoveryListener$1@61ea2cff] [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：本地部署的部署元数据：GridDeploymentMetadata [depMode=SHARED，别名=org。 apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1, clsName=org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1, userVer=null, sndNodeId=2e573c60-45f0-4429-a3fa-068489663148, clsLdrId= null，clsLdr=null，参与者=null，parentLdr=null，record=true，nodeFilter=null，seqNum=n/a] [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：注册 [ldrRsrcs={org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f ={org.apache.ignite.internal.util.typedef.T2=org.apache.ignite.internal.util.typedef.T2, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap= org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap，org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionFullMap=org.apache.ignite.internal.processors。 cache.distributed.dht.preloader.GridDhtPartitionFullMap，java.util.Collections$UnmodifiableList=java.util.Collections$UnmodifiableList，org.apache.ignite.internal.visor.cache.VisorCacheMetricsCollectorTask=org.apache.ignite.internal.visor。 cache.VisorCacheMetricsCollectorTask, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMap =org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMap, org.apache.ignite.internal.processors.service.GridServiceProcessor$1=org.apache.ignite.internal.processors.service.GridServiceProcessor$1 , org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionMap=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionMap, org.apache.ignite.internal.processors .cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap, org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1=org. apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1}}, ldr=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f, rsrc=class org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1] [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：要注册的资源：{org.apache.ignite.internal.processors.datastructures .DataStructuresProcessor$1$1=org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1} [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：新资源：{org.apache.ignite.internal.processors.datastructures。 DataStructuresProcessor$1$1=org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1} [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- LocalDeploymentSpi：删除资源 [clsLdrToIgnore=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f , rsrcs={org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1=org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1}] [2017-10-09 14:26:52.211] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：从 spi 检索自动加载的资源：DeploymentResourceAdapter [name=org.apache。 ignite.internal.processors.datastructures.DataStructuresProcessor$1$1, rsrcCls=class org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1, clsLdr=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f] [2017-10-09 14:26:52.212] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridDeploymentLocalStore：获取的部署类：GridDeployment [ts=1507576972855，depMode=SHARED，clsLdr= org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f, clsLdrId=6d9e6920f51-2e573c60-45f0-4429-a3fa-068489663148, userVer=0, loc=true, sampleClsName=org.apache.ignite.internal.processors.cache.distributed。 dht.preloader.GridDhtPartitionFullMap，pendingUndeploy=false，undeployed=false，usage=0] [2017-10-09 14:26:52.212] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- GridResourceProcessor：注入资源 [target=org.apache.ignite.internal.processors.datastructures .DataStructuresProcessor$1$1@1ff41d49] [2017-10-09 14:26:52.213] 启动 - 9955 调试 [pub-#34%null%] --- GridClosureProcessor：网格可运行开始：closure-proc-worker [2017-10-09 14:26:52.213] 启动 - 9955 调试 [pub-#34%null%] --- GridClosureProcessor：网格可运行正常完成：closure-proc-worker [2017-10-09 14:26:52.216] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- TcpCommunicationSpi：由于节点已离开，强制 NIO 客户端关闭 [nodeId=4cc6c321-d9cc-4149 -a6ef-cba68877a269, client=GridTcpNioCommunicationClient [ses=GridSelectorNioSessionImpl [worker=DirectNioClientWorker [super=AbstractNioClientWorker [idx=3, bytesRcvd=714, bytesSent=6799, bytesRcvd0=0, bytesSent0=0, select=true, super=GridWorker [name =grid-nio-worker-tcp-comm-3，igniteInstanceName=null，finished=false，hashCode=1557230104，interrupted=false，runner=grid-nio-worker-tcp-comm-3-#20%null%]] ], writeBuf=java.nio.DirectByteBuffer[pos=0 lim=32768 cap=32768], readBuf=java.nio.DirectByteBuffer[pos=0 lim=32768 cap=32768], inRecovery=GridNioRecoveryDescriptor [acked=0, resendCnt= 0, rcvCnt=1, sentCnt=1, reserved=true, lastAck=0, nodeLeft=false, node=TcpDiscoveryNode [id=4cc6c321-d9cc-4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17 .0.1], sockAddrs=[/172.17.0.1:0, /127.0.0.1:0, /10.70.255.8:0], discPo rt=0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], connected=true, connectCnt=0, queueLimit=4096, reserveCnt =1，pairedConnections=false]，outRecovery=GridNioRecoveryDescriptor [acked=0，resendCnt=0，rcvCnt=1，sentCnt=1，reserved=true，lastAck=0，nodeLeft=false，node=TcpDiscoveryNode [id=4cc6c321-d9cc- 4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17.0.1:0, /127.0.0.1:0, /10.70.255.8:0], discPort=0 , order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], connected=true, connectCnt=0, queueLimit=4096, reserveCnt=1, pairConnections=false], super=GridNioSessionImpl [locAddr=/10.70.242.138:47100, rmtAddr=/10.70.255.8:53916, createTime=1507577162587, closeTime=0, bytesSent=6799, bytesRcvd=714, bytesSent0=0, bytesRcvd0=0 , sndSchedTime=1507577162587, lastSndTime=1507577162697, lastRcvTime=1507577162617, readsPaused=false, 过滤erChain=FilterChain[filters=[GridNioCodecFilter [parser=org.apache.ignite.internal.util.nio.GridDirectParser@9573b3b, directMode=true], GridConnectionBytesVerifyFilter], 接受=true]], super=GridAbstractCommunicationClient [lastUsed=1507577162587, 关闭=假，connIdx=0]]] [2017-10-09 14:26:52.217] 启动 - 9955 调试 [disco-event-worker-#26%null%] --- TcpCommunicationSpi：提供移动 [ses=GridSelectorNioSessionImpl [worker=DirectNioClientWorker [super=AbstractNioClientWorker [idx] =3，bytesRcvd=714，bytesSent=6799，bytesRcvd0=0，bytesSent0=0，select=true，super=GridWorker [name=grid-nio-worker-tcp-comm-3，igniteInstanceName=null，finished=false，hashCode =1557230104, 中断=false, runner=grid-nio-worker-tcp-comm-3-#20%null%]]], writeBuf=java.nio.DirectByteBuffer[pos=0 lim=32768 cap=32768], readBuf =java.nio.DirectByteBuffer[pos=0 lim=32768 cap=32768], inRecovery=GridNioRecoveryDescriptor [acked=0, resendCnt=0, rcvCnt=1, sentCnt=1, reserved=true, lastAck=0, nodeLeft=false, node=TcpDiscoveryNode [id=4cc6c321-d9cc-4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17.0.1:0, /127.0.0.1:0, / 10.70.255.8:0], discPort=0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient= true]，connected=true，connectCnt=0，queueLimit=4096，reserveCnt=1，pairedConnections=false]，outRecovery=GridNioRecoveryDescriptor [acked=0，resendCnt=0，rcvCnt=1，sentCnt=1，reserved=true，lastAck= 0, nodeLeft=false, node=TcpDiscoveryNode [id=4cc6c321-d9cc-4149-a6ef-cba68877a269, addrs=[10.70.255.8, 127.0.0.1, 172.17.0.1], sockAddrs=[/172.17.0.1:0, /127.0 .0.1:0, /10.70.255.8:0], discPort=0, order=57, intOrder=31, lastExchangeTime=1507577126368, loc=false, ver=2.1.0#20170720-sha1:a6ca5c8a, isClient=true], connected=true，connectCnt=0，queueLimit=4096，reserveCnt=1，pairedConnections=false]，super=GridNioSessionImpl [locAddr=/10.70.242.138:47100，rmtAddr=/10.70.255.8:53916，createTime=1507577162587，closeTime=0 , bytesSent=6799, bytesRcvd=714, bytesSent0=0, bytesRcvd0=0, sndSchedTime=1507577162587, lastSndTime=1507577162697, lastRcvTime=1507577162617, readsPaused=false, filterChain=FilterChain[filters=[GridNioCodecFilter [parser=org.apache. internal.util.nio.GridDirectParser@9573b3 b、directMode=true]、GridConnectionBytesVerifyFilter]、accepted=true]]、fut=NioOperationFuture [op=CLOSE]] [2017-10-09 14:26:52.217] boot - 9955 DEBUG [disco-event-worker-#26%null%] --- GridIoManager：从发现启动延迟列表中删除消息（发送方节点左侧拓扑）：null [2017-10-09 14:26:52.217] 启动 - 9955 调试 [pub-#35%null%] --- GridClosureProcessor：网格可运行开始：closure-proc-worker [2017-10-09 14:26:52.217] 引导 - 9955 调试 [pub-#35%null%] --- GridClosureProcessor：网格可运行正常完成：closure-proc-worker [2017-10-09 14:26:52.283] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- GridTimeoutProcessor：发生超时：org.apache.ignite.internal.processors.cache。 transactions.IgniteTxManager$NodeFailureTimeoutObject@7ff59c90 [2017-10-09 14:26:52.284] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- GridDeploymentLocalStore：本地部署的部署元数据：GridDeploymentMetadata [depMode=SHARED，别名=org。 apache.ignite.internal.processors.cache.transactions.IgniteTxManager$NodeFailureTimeoutObject$2, clsName=org.apache.ignite.internal.processors.cache.transactions.IgniteTxManager$NodeFailureTimeoutObject$2, userVer=null, sndNodeId=2e573c60-45f0-4429- a3fa-068489663148，clsLdrId=null，clsLdr=null，参与者=null，parentLdr=null，record=true，nodeFilter=null，seqNum=n/a] [2017-10-09 14:26:52.285] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- LocalDeploymentSpi：注册 [ldrRsrcs={org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f ={org.apache.ignite.internal.util.typedef.T2=org.apache.ignite.internal.util.typedef.T2, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap= org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionHistorySuppliersMap，org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionFullMap=org.apache.ignite.internal.processors。 cache.distributed.dht.preloader.GridDhtPartitionFullMap，java.util.Collections$UnmodifiableList=java.util.Collections$UnmodifiableList，org.apache.ignite.internal.visor.cache.VisorCacheMetricsCollectorTask=org.apache.ignite.internal.visor。 cache.VisorCacheMetricsCollectorTask, org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMa p=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionsToReloadMap, org.apache.ignite.internal.processors.service.GridServiceProcessor$1=org.apache.ignite.internal.processors.service.GridServiceProcessor $1, org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1=org.apache.ignite.internal.processors.datastructures.DataStructuresProcessor$1$1, org.apache.ignite.internal.processors.cache.distributed.dht。 preloader.GridDhtPartitionMap=org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPartitionMap，org.apache.ignite.internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap=org.apache.ignite。 internal.processors.cache.distributed.dht.preloader.IgniteDhtPartitionCountersMap，org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1=org.apache.ignite.internal.processors.task.GridTaskProcessor$TaskDiscoveryListener$1}}， ldr=org.springframework.boot.loader.Lau nchedURLClassLoader@7adf9f5f, rsrc=class org.apache.ignite.internal.processors.cache.transactions.IgniteTxManager$NodeFailureTimeoutObject$2] 2017-10-09 14:26:52.285] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- GridDeploymentLocalStore：从 spi 检索自动加载的资源：DeploymentResourceAdapter [name=org.apache.ignite .internal.processors.cache.transactions.IgniteTxManager$NodeFailureTimeoutObject$2, rsrcCls=class org.apache.ignite.internal.processors.cache.transactions.IgniteTxManager$NodeFailureTimeoutObject$2, clsLdr=org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f] [2017-10-09 14:26:52.285] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- GridDeploymentLocalStore：获取的部署类：GridDeployment [ts=1507576972855，depMode=SHARED，clsLdr= org.springframework.boot.loader.LaunchedURLClassLoader@7adf9f5f, clsLdrId=6d9e6920f51-2e573c60-45f0-4429-a3fa-068489663148, userVer=0, loc=true, sampleClsName=org.apache.ignite.internal.processors.cache.distributed。 dht.preloader.GridDhtPartitionFullMap，pendingUndeploy=false，undeployed=false，usage=0] [2017-10-09 14:26:52.285] 启动 - 9955 调试 [grid-timeout-worker-#15%null%] --- GridResourceProcessor：注入资源 [target=org.apache.ignite.internal.processors.cache .transactions.IgniteTxManager$NodeFailureTimeoutObject$2@3f183e4] [2017-10-09 14:26:52.317] 启动 - 9955 调试 [http-nio-8081-exec-8] --- CacheHelper：访问缓存 ng-security-service-ORG_SPEC_CACHE 的总时间 * |值 com.cache.model.PrefixCluster@6954be5d：167 毫秒 [2017-10-09 14:26:52.319] 启动 - 9955 调试 [http-nio-8081-exec-8] --- OrgSpecCacheImpl：OrgSpec 缓存统计信息：OrgSpec ObjId：IgniteCacheProxy [delegate=GridNearCacheAdapter []，opCtx=null , restartFut=null] HitCount: 126, MissCount: 53, AvgReadTime: 126, Eviction Count: 0 [2017-10-09 14:26:52.321] 启动 - 9955 调试 [sys-#36%null%] --- GridClosureProcessor：网格可运行开始：closure-proc-worker

我的问题是，这是预期的行为吗？我们能否让近端缓存不被绕过，或者至少在坏客户端断开连接后使用近端缓存重新建立。

Answer 1

事实证明它们是近缓存的错误。当拓扑发生变化时，它可以清除 NearCacheGridEntry 上的拓扑版本，这将导致每次调用检查条目是否有效返回 false。

Bug 已提交https://issues.apache.org/jira/browse/IGNITE-6767