C# 客户端节点无法连接到服务器

我使用的是Ignite.NET 2.7.6，我有一个服务器节点和多个客户端节点。经过一段时间（几天）后，一些客户端节点无法连接到服务器节点。唯一有帮助的是重新启动服务器

好的，如果是连接问题，那么重启就不起作用了，但是它有帮助，所以我不明白出了什么问题

以下是来自客户端断开连接的服务器节点的日志：

[23:13:17,028][INFO][tcp-disco-srvr-#3][TcpDiscoverySpi] TCP discovery accepted incoming connection [rmtAddr=/192.168.22.35, rmtPort=61608]
[23:13:17,028][INFO][tcp-disco-srvr-#3][TcpDiscoverySpi] TCP discovery spawning a new thread for connection [rmtAddr=/192.168.22.35, rmtPort=61608]
[23:13:17,029][INFO][tcp-disco-sock-reader-#1093][TcpDiscoverySpi] Started serving remote node connection [rmtAddr=/192.168.22.35:61608, rmtPort=61608]
[23:13:17,034][INFO][tcp-disco-sock-reader-#1093][TcpDiscoverySpi] Finished serving remote node connection [rmtAddr=/192.168.22.35:61608, rmtPort=61608
[23:13:17,437][INFO][tcp-disco-srvr-#3][TcpDiscoverySpi] TCP discovery accepted incoming connection [rmtAddr=/192.168.22.35, rmtPort=61630]
[23:13:17,437][INFO][tcp-disco-srvr-#3][TcpDiscoverySpi] TCP discovery spawning a new thread for connection [rmtAddr=/192.168.22.35, rmtPort=61630]
[23:13:17,438][INFO][tcp-disco-sock-reader-#1095][TcpDiscoverySpi] Started serving remote node connection [rmtAddr=/192.168.22.35:61630, rmtPort=61630]
[23:13:17,467][INFO][disco-event-worker-#72][GridDiscoveryManager] Added new node to topology: TcpDiscoveryNode [id=40bdc648-b25f-4b16-ac39-013bd7390416, addrs=[127.0.0.1, 192.168.22.35], sockAddrs=[/127.0.0.1:0, BB-REC-VIP.ast.loc/192.168.22.35:0], discPort=0, order=477, intOrder=257, lastExchangeTime=1581613997440, loc=false, ver=2.7.6#20190911-sha1:21f7ca41, isClient=true]
[23:13:17,468][INFO][disco-event-worker-#72][GridDiscoveryManager] Topology snapshot [ver=477, locNode=4fdcbe35, servers=1, clients=36, state=ACTIVE, CPUs=162, offheap=16.0GB, heap=27.0GB]
[23:13:17,468][INFO][disco-event-worker-#72][GridDiscoveryManager]   ^-- Baseline [id=3, size=4, online=1, offline=3]
[23:13:17,468][INFO][exchange-worker-#73][time] Started exchange init [topVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], mvccCrd=MvccCoordinator [nodeId=4fdcbe35-8390-4584-92a4-8d918625540e, crdVer=1581589948994, topVer=AffinityTopologyVersion [topVer=1, minorTopVer=0]], mvccCrdChange=false, crd=true, evt=NODE_JOINED, evtNode=40bdc648-b25f-4b16-ac39-013bd7390416, customEvt=null, allowMerge=true]
[23:13:17,469][INFO][exchange-worker-#73][GridDhtPartitionsExchangeFuture] Finish exchange future [startVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], resVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], err=null]
[23:13:17,470][INFO][exchange-worker-#73][GridDhtPartitionsExchangeFuture] Completed partition exchange [localNode=4fdcbe35-8390-4584-92a4-8d918625540e, exchange=GridDhtPartitionsExchangeFuture [topVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], evt=NODE_JOINED, evtNode=TcpDiscoveryNode [id=40bdc648-b25f-4b16-ac39-013bd7390416, addrs=[127.0.0.1, 192.168.22.35], sockAddrs=[/127.0.0.1:0, BB-REC-VIP.ast.loc/192.168.22.35:0], discPort=0, order=477, intOrder=257, lastExchangeTime=1581613997440, loc=false, ver=2.7.6#20190911-sha1:21f7ca41, isClient=true], done=true], topVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], durationFromInit=0]
[23:13:17,470][INFO][exchange-worker-#73][time] Finished exchange init [topVer=AffinityTopologyVersion [topVer=477, minorTopVer=0], crd=true]
[23:13:17,489][INFO][exchange-worker-#73][GridCachePartitionExchangeManager] Skipping rebalancing (nothing scheduled) [top=AffinityTopologyVersion [topVer=477, minorTopVer=0], force=false, evt=NODE_JOINED, node=40bdc648-b25f-4b16-ac39-013bd7390416]
[23:13:37,550][WARNING][grid-timeout-worker-#47][TcpDiscoverySpi] Socket write has timed out (consider increasing 'sockTimeout' configuration property) [sockTimeout=20000, rmtAddr=/192.168.22.35:61630, rmtPort=61630, sockTimeout=20000]
[23:13:37,550][INFO][tcp-disco-sock-reader-#1095][TcpDiscoverySpi] Finished serving remote node connection [rmtAddr=/192.168.22.35:61630, rmtPort=61630

---

有没有解决此问题的建议？

从提供的日志中，我看到客户端已连接。在客户端断开连接的地方提供日志会很好。完整的服务器日志也很有用。顺便说一句，您提到了一个节点集群，但我看到您的基线包含4个基线[id=3，size=4，online=1，offline=3]。某些分区位于脱机节点上，这可能是一个问题。您可以阅读。服务器处于复制模式，所有脱机服务器都处于脱机状态很长一段时间，这是一个结束的实验。关于“客户端断开连接的位置”：我犯了错误-这是服务器节点的日志，该客户端断开连接的位置“[23:13:17467][INFO][disco event worker-#72][GridDiscoveryManager]在拓扑中添加了新节点：TcpDiscoveryNode[id=40bdc648-b25f-4b16-ac39-013bd7390416，addrs=[127.0.0.1，192.168.22.35]，sockAddrs=[/127.0.0.1:0，BB-REC-VIP.ast.loc/192.168.22.35:0]，disport=0，order=477，intOrder=257，lastExchangeTime=1581613997440，loc=false，ver=2.7.6#20190911-sha1:21f7ca41，isClient=true]“-这意味着客户端已连接到集群。（isClient=true）和（向拓扑中添加了新节点）消息显示。请重新检查您的日志。它应该包含服务器日志中的节点\u LEFT或节点\u FAILED消息。“已完成远程节点连接服务”这并不意味着客户端的连接已关闭？我将再次检查下一次连接何时发生，谢谢。这不是真的。这只是意味着某些TCP连接已关闭。例如，一个节点向另一个节点请求度量或某些任务结果。收到结果后，可以关闭连接。您应该调查日志服务器将在其中写入某个节点已断开连接或离开群集。它还应生成拓扑更改（拓扑快照消息）。请尝试为节点\u离开或节点\u失败的消息刷新服务器日志，并提供它们。