应用重启不完全问题

[chenchongqi]

昨天239.20通过更新系统更新重启完后，jdk-out日志中一直有以下输出：

[11-10 17:14:15.975] {main} [PcRoute] use dns server: system default
[11-10 17:14:16.013] {main} [PcRoute] md5 not changed, do not need reload
[11-10 17:14:16.056] {main} [PcRoute] reload vServerPool OK.
[11-10 17:14:16.242] {main} WebApp[http://localhost:8081] Initializing Spring root WebApplicationContext
[11-10 17:14:17.413] {main} Unknown or Unsupported Option 'itpricelibindex'
[11-10 17:14:17.463] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.111:5506
[11-10 17:14:17.468] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.112:5506
[11-10 17:14:17.469] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.117:5506
[11-10 17:14:17.471] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.83:5506
[11-10 17:14:18.806] {main} Unknown or Unsupported Option 'itpricelib'
[11-10 17:14:18.813] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.82:5506
[11-10 17:14:18.817] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.81:5506
[11-10 17:14:18.839] {ReplicaSetStatus:Updater} Server seen up: 192.168.237.92:5506
[11-10 17:14:19.144] {main} WebApp[http://localhost:8081] active
[11-10 17:14:19.168] {main} WebApp[http://localhost:8081/robin_admin] active
[11-10 17:14:19.169] {main} Host[] active
[11-10 17:14:19.172] {main} Loaded Socket JNI library.
[11-10 17:14:24.295] {main} java.net.BindException: Address already in use
[11-10 17:14:24.295] {main} Can't bind to /127.0.0.1:6800.
[11-10 17:14:24.295] {main} Check for another server listening to that port.
[11-10 17:14:24.295] {main}     at com.caucho.vfs.QJniServerSocket.create(QJniServerSocket.java:103)
[11-10 17:14:24.295] {main}     at com.caucho.server.port.Port.bind(Port.java:964)
[11-10 17:14:24.295] {main}     at com.caucho.server.cluster.Server.bindPorts(Server.java:1379)
[11-10 17:14:24.295] {main}     at com.caucho.server.cluster.Server.start(Server.java:1325)
[11-10 17:14:24.295] {main}     at com.caucho.server.cluster.Cluster.startServer(Cluster.java:710)
[11-10 17:14:24.295] {main}     at com.caucho.server.cluster.ClusterServer.startServer(ClusterServer.java:542)
[11-10 17:14:24.295] {main}     at com.caucho.server.resin.Resin.start(Resin.java:703)
[11-10 17:14:24.295] {main}     at com.caucho.server.resin.Resin.initMain(Resin.java:1162)
[11-10 17:14:24.295] {main}     at com.caucho.server.resin.Resin.main(Resin.java:1365)
[11-10 17:14:24.296] {main} Server[id=app-a,cluster=app-tier] error
[11-10 17:14:27.027] {main} Proxy Cache disk-size=1024M memory-size=64M
[11-10 17:14:27.040] {main} PingThread[] starting, checking []
[11-10 17:14:27.160] {main}
[11-10 17:14:27.161] {main} Linux 2.6.9-67.ELsmp amd64
[11-10 17:14:27.161] {main} Java(TM) SE Runtime Environment 1.6.0_06-b02, GBK, zh
[11-10 17:14:27.161] {main} Java HotSpot(TM) 64-Bit Server VM 10.0-b22, 64, mixed mode, Sun Microsystems Inc.
[11-10 17:14:27.161] {main} user.name: root
[11-10 17:14:27.161] {main} resin.home = /data/PRG/resin-pro-3.1.8
[11-10 17:14:27.161] {main} resin.root = /data/PRG/resin-pro-3.1.8
[11-10 17:14:27.161] {main} resin.conf = /data/PRG/resin-pro-3.1.8/conf/resin.conf.product.xml

然后无法自己恢复，前台访问正常，但是从日志看应用在不断初始化一些资源，对mc和数据库会有一定压力，以后还是要加强更新后的日志检查来避免。

林浩(林浩#gz4046) 11:39:02
看了一下log，应该是resin的watchdog干的，watchdog进程发现有异常就重启，但6600 6800没杀掉，就这样
这个再观察看看吧
林浩(林浩#gz4046) 11:39:15
应该不会太常出现的