分布式索引系统配置文件<设想一>

索引模式(稳定,快速,可迅速恢同步,可迅速恢复性,同一结点间的数据也能讯速同步\)
0\适合对数据库
1\能正常索引数据
2\能把索引的索引数据增量传送给集群结点
3\当总索引系统出错时,能经过重启动,迅速恢复正常,并能把数据适时同步到集群结点
4\当集群结点主服务器出现异常时,能迅速探测到该结点中其他机器的可用性,并且临时指定该机器为该结点中主服务器,索引服务器要记录这种情况的异常情况,当原来主服器恢复后,索引数据能同步


<?xml version="1.0" encoding="UTF-8"?>
<dichan.com>
<Index.search-ref>
<businessid>1</businessid>
<description>架构上:一个满足分布式搜索的索引系统(假定有三个(可以线性增加)结点(每个结点也有三台(可以线性增加)机器组成,作为冗余)组成的一个分布搜索系统)</description>
<!--2*60*1000 ms-->
<intervaltime>120000</intervaltime>
<recordnum>1000</recordnum>
<cyclecount>20</cyclecount>
<!-- 单位：ms -->
<sleeptime>10000</sleeptime>
<!--取不到数时，重试增量时间，单位:2*24*60*60*1000ms 两天 -->
<increatment>172800000</increatment>
<!--删除多少天前的数据，单位:小时 default 10*24 220-->
<ndaybeferorenotetable>220</ndaybeferorenotetable>
<clusters>
<description>一个集群结点中允许多台服务器执行相同功能,并且必须指定一台主服务器</description>
<cluster>
<name>分布式搜索集群结点一</name>
<node>1</node>
<!--是否启用-->
<use>1</use>
<addresss>
<address>
<seq>1</seq>
<ip>192.168.1.1</ip>
<port>8089</port>
<master>1</master>
<weight>1</weight>
</address>
<address>
<seq>12</seq>
<ip>192.168.1.2</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
<address>
<seq>3</seq>
<ip>192.168.1.3</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
</addresss>
<!--数据是一个大概范围,如 0,0 表示不限ID范围-->
<idrange>0,10000000</idrange>
<!--可对idrange范围的值进行拆分,以加快索引速度-->
<shard>10</shard>

<!--增量索引时,时间开始点,如果是0,表示从数据库中ID从小到大开始-->
<indexstarttime_update>2010-06-09 14:06:11.387</indexstarttime_update>
<!--错开的时间10*60*1000-->
<indexmerge>600000</indexmerge>
<!--数据库连接相关-->
<dbdata.business>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.business>
<dbdata.indexsystem>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.indexsystem>
<dbdata.log>
<dbid>17</dbid>
<dbname>日志相关</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=数据库;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>16</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.log>
<dichanIndex>
<businessseq>1</businessseq>
<businessname>业务名</businessname>
<businessindexpath>F:\file\index\业务名代号\结点\业务表\index</businessindexpath>
<businesslogfilename>F:\file\index\业务名代号\结点\业务表\Log\index.txt</businesslogfilename>
<businesslogpath>F:\file\index\业务名代号\结点\业务表\Log</businesslogpath>
</dichanIndex>
</cluster>
<cluster>
<name>分布式搜索集群结点二</name>
<node>2</node>
<!--是否启用-->
<use>1</use>
<addresss>
<address>
<seq>1</seq>
<ip>192.168.1.21</ip>
<port>8089</port>
<master>1</master>
<weight>1</weight>
</address>
<address>
<seq>2</seq>
<ip>192.168.1.22</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
<address>
<seq>3</seq>
<ip>192.168.1.23</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
</addresss>
<idrange>10000000,20000000</idrange>
<!--可对idrange范围的值进行拆分,以加快索引速度-->
<shard>10</shard>

<!--增量索引时,时间开始点,如果是0,表示从数据库中ID从小到大开始-->
<indexstarttime_update>2010-06-09 14:06:11.387</indexstarttime_update>
<!--数据库连接相关-->
<dbdata.business>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.business>
<dbdata.indexsystem>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.indexsystem>
<dbdata.log>
<dbid>17</dbid>
<dbname>日志相关</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=数据库;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>16</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.log>
<dichanIndex>
<businessseq>1</businessseq>
<businessname>业务名</businessname>
<businessindexpath>F:\file\index\业务名代号\结点\业务表\index</businessindexpath>
<businesslogfilename>F:\file\index\业务名代号\结点\业务表\Log\index.txt</businesslogfilename>
<businesslogpath>F:\file\index\业务名代号\结点\业务表\Log</businesslogpath>
</dichanIndex>

</cluster>

<cluster>
<name>分布式搜索集群结点三</name>
<node>3</node>
<!--是否启用-->
<use>1</use>
<addresss>
<address>
<seq>1</seq>
<ip>192.168.1.41</ip>
<port>8089</port>
<master>1</master>
<weight>1</weight>
</address>
<address>
<seq>2</seq>
<ip>192.168.1.42</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
<address>
<seq>3</seq>
<ip>192.168.1.43</ip>
<port>8089</port>
<master>0</master>
<weight>1</weight>
</address>
</addresss>
<idrange>20000000,30000000</idrange>
<!--可对idrange范围的值进行拆分,以加快索引速度-->
<shard>10</shard>

<!--增量索引时,时间开始点,如果是0,表示从数据库中ID从小到大开始-->
<indexstarttime_update>2010-06-09 14:06:11.387</indexstarttime_update>
<!--数据库连接相关-->
<dbdata.business>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.business>
<dbdata.indexsystem>
<dbid>1</dbid>
<dbname>测试</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=xxxxxx;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>14</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.indexsystem>
<dbdata.log>
<dbid>17</dbid>
<dbname>日志相关</dbname>
<serverip>10.1.3.252</serverip>
<db-driver>com.microsoft.sqlserver.jdbc.SQLServerDriver</db-driver>
<db-url>jdbc:sqlserver://10.1.3.252;databaseName=数据库;SelectMethod=cursor</db-url>
<set-tran-level>false</set-tran-level>
<select-tran-level>1</select-tran-level>
<update-tran-level>2</update-tran-level>
<max-connections>16</max-connections>
<db-user>xxx</db-user>
<db-password>xxx</db-password>
</dbdata.log>
<dichanIndex>
<businessseq>1</businessseq>
<businessname>业务名</businessname>
<businessindexpath>F:\file\index\业务名代号\结点\业务表\index</businessindexpath>
<businesslogfilename>F:\file\index\业务名代号\结点\业务表\Log\index.txt</businesslogfilename>
<businesslogpath>F:\file\index\业务名代号\结点\业务表\Log</businesslogpath>
</dichanIndex>
</cluster>

</clusters>
<Index.search-ref>
</dichan.com>


监控模式
0\适时找出总索引与集群结点的索引数据的情况