MongoDB分片集群shard节点宕机后的处理
版权声明：本文为博主原创文章，转载时需要带上原文链接。https://blog.csdn.net/weixin_39032311/article/details/92805072
方案：用新shard节点替换宕掉的shard节点
一、新建shard节点并启动（假设新shard节点的端口号为27028）
二、在mongos上删除原shard节点的配置信息
> use admin
> db.runCommand({removeShard:"10.201.81.105:27018"})
可能出现以下错误信息：
mongos> db.runCommand({removeShard:"10.201.81.105:27018"})
{
"msg" : "draining ongoing",
"state" : "ongoing",
"remaining" : {
"chunks" : NumberLong(0),
"dbs" : NumberLong(1)
},
"note" : "you need to drop or movePrimary these databases",
"dbsToMove" : [
"mgotest2"
],
"ok" : 1,
"operationTime" : Timestamp(1560822195, 1),
"$clusterTime" : {
"clusterTime" : Timestamp(1560822195, 1),
"signature" : {
"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
"keyId" : NumberLong(0)
}
}
}
注意：如果删除的片是数据库的大本营(基片)，必须手动移动或删除数据库，用moveprimary命令，上面的示例中就提示10.201.81.105:27018是mgotest2库的大本营(primary)，这个信息可以通过查看config.databases看到：
mongos> use config
switched to db config
mongos> db.databases.find(）
{ "_id" : "mgotest2", "primary" : "shardsvr2", "partitioned" : false, "version" : { "uuid" : UUID("360e1a26-b6f9-44a5-90ea-a148bf854e59"), "lastMod" : 1 } }
此操作需要主节点10.201.81.105:27018正常运行状态下进行，否则会出现以下错误
mongos> use admin
switched to db admin
mongos> db.runCommand({"moveprimary":"mgotest2","to":"10.201.81.218:27018"})
{
"ok" : 0,
"errmsg" : "Could not find host matching read preference { mode: \"primary\" } for set shardsvr2",
"code" : 133,
"codeName" : "FailedToSatisfyReadPreference",
"operationTime" : Timestamp(1560823671, 1),
"$clusterTime" : {
"clusterTime" : Timestamp(1560823671, 1),
"signature" : {
"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
"keyId" : NumberLong(0)
}
}
}
若主节点10.201.81.105:27018不能正常运行，则进行以下操作
mongos> use config
switched to db config
mongos> db.shards.find()
{ "_id" : "shardsvr1", "host" : "shardsvr1/10.201.81.218:27018", "state" : 1 }
{ "_id" : "shardsvr2", "host" : "shardsvr2/10.201.81.105:27018", "state" : 1, "draining" : true }
mongos> db.databases.find()
{ "_id" : "mgotest2", "primary" : "shardsvr2", "partitioned" : false, "version" : { "uuid" : UUID("360e1a26-b6f9-44a5-90ea-a148bf854e59"), "lastMod" : 1 } }
mongos> db.databases.remove({"_id" : "mgotest2"})
WriteResult({ "nRemoved" : 1 })
进行以上步骤后，可以正常删除原shard节点的配置信息
三、添加新shard节点的配置信息
mongos> sh.addShard("shardsvr1/10.201.81.105:27028")
{
"shardAdded" : "shardsvr1",
"ok" : 1,
"operationTime" : Timestamp(1560825183, 1),
"$clusterTime" : {
"clusterTime" : Timestamp(1560825183, 1),
"signature" : {
"hash" : BinData(0,"AAAAAAAAAAAAAAAAAAAAAAAAAAA="),
"keyId" : NumberLong(0)
}
}
}