NOSQL 之 cassadra 安装与集群配置

cassadra 安装

1.下载：

apache-cassandra-2.0.6-bin.tar.gz

jdk-8-linux-x64.rpm

2.先更新jdk

rpm -ivh jdk-8-linux-x64.rpm

修改java ，指定到最新版本

rm /usr/bin/java java.bak
ln -s /usr/java/jdk1.8.0/bin/java java

[root@localhost bin]# java -version
   java version "1.8.0"
   Java(TM) SE Runtime Environment (build 1.8.0-b132)
   Java HotSpot(TM) 64-Bit Server VM (build 25.0-b70, mixed mode)

3.安装cassandra

[root@localhost backup]# tar zxvf apache-cassandra-2.0.6-bin.tar.gz

[root@localhost backup]# cd apache-cassandra-2.0.6/bin

root@localhost bin]# ll cassandra
-rwxr-xr-x 1 mongo mongo 7108 3月 7 20:34 cassandra
[root@localhost bin]# ./cassndra
-bash: ./cassndra: 没有那个文件或目录
[root@localhost bin]# ./cassandra
[root@localhost bin]# INFO 17:59:33,186 Logging initialized
INFO 17:59:33,369 Loading settings from file:/opt/backup/apache-cassandra-2.0.6/conf/cassandra.yaml
ps INFO 17:59:34,440 Data files directories: [/var/lib/cassandra/data]
INFO 17:59:34,443 Commit log directory: /var/lib/cassandra/commitlog
INFO 17:59:34,444 DiskAccessMode 'auto' determined to be mmap, indexAccessMode is mmap

Cassandra集群实施方案：

两个节点：192.168.10.104
192.168.10.206

相关的JAVA，cassandra 安装不再累述。两节点的配置安装和上面的是一样的。
下面来做相关配置文件的修改：

节点1：

192.168.10.104

vi /opt/apache-cassandra-2.0.6/conf/cassandra.yaml

修改内容如下：

cluster_name: 'Cluster1'

- seeds: "192.168.10.104"

listen_address: 192.168.10.104

节点2：

192.168.10.206

vi /opt/apache-cassandra-2.0.6/conf/cassandra.yaml

修改内容如下：

cluster_name: 'Cluster1'

- seeds: "192.168.10.104"

listen_address: 192.168.10.206

启动

节点1：

[root@localhost bin]# ./cassandra -f
INFO 21:30:13,864 Logging initialized
INFO 21:30:13,907 Loading settings from file:/opt/backup/apache-cassandra-2.0.6/conf/cassandra.yaml
INFO 21:30:14,202 Data files directories: [/var/lib/cassandra/data]
INFO 21:30:14,203 Commit log directory: /var/lib/cassandra/commitlog
INFO 21:30:14,203 DiskAccessMode 'auto' determined to be mmap, indexAccessMode is mmap
INFO 21:30:14,203 disk_failure_policy is stop
INFO 21:30:14,203 commit_failure_policy is stop
INFO 21:30:14,207 Global memtable threshold is enabled at 1996MB
INFO 21:30:14,282 Not using multi-threaded compaction
INFO 21:30:14,420 JVM vendor/version: Java HotSpot(TM) 64-Bit Server VM/1.8.0
INFO 21:30:14,420 Heap size: 8375238656/8375238656

...................

INFO 22:14:41,166 Compacted 4 sstables to [/var/lib/cassandra/data/system/local/system-local-jb-25,]. 6,206 bytes to 5,671 (~91% of original) in 26ms = 0.208011MB/s. 4 total partitions merged to 1. Partition merge counts were {4:1, }
INFO 22:14:41,197 Node /192.168.10.104 state jump to normal
INFO 22:14:41,209 Waiting for gossip to settle before accepting client requests...
INFO 22:14:49,210 No gossip backlog; proceeding
INFO 22:14:49,351 Starting listening for CQL clients on /192.168.10.104:9042...
INFO 22:14:49,406 Using TFramedTransport with a max frame size of 15728640 bytes.
INFO 22:14:49,408 Binding thrift service to /192.168.10.104:9160
INFO 22:14:49,418 Using synchronous/threadpool thrift server on 192.168.10.104 : 9160
INFO 22:14:49,419 Listening for thrift clients...
INFO 22:15:44,802 Handshaking version with /192.168.10.206
INFO 22:15:45,914 Node /192.168.10.206 is now part of the cluster
INFO 22:15:45,924 Handshaking version with /192.168.10.206
INFO 22:15:45,930 InetAddress /192.168.10.206 is now UP

节点2：

[root@localhost bin]# ./cassandra -f
INFO 22:18:50,672 Logging initialized
INFO 22:18:50,742 Loading settings from file:/opt/backup/apache-cassandra-2.0.6/conf/cassandra.yaml
INFO 22:18:51,194 Data files directories: [/var/lib/cassandra/data]
INFO 22:18:51,196 Commit log directory: /var/lib/cassandra/commitlog
INFO 22:18:51,196 DiskAccessMode 'auto' determined to be mmap, indexAccessMode is mmap
INFO 22:18:51,196 disk_failure_policy is stop
INFO 22:18:51,197 commit_failure_policy is stop
INFO 22:18:51,203 Global memtable threshold is enabled at 253MB
INFO 22:18:51,339 Not using multi-threaded compaction
INFO 22:18:51,651 JVM vendor/version: Java HotSpot(TM) 64-Bit Server VM/1.8.0
INFO 22:18:51,652 Heap size: 1063256064/1063256064
...................

INFO 22:19:05,145 Starting listening for CQL clients on /192.168.10.206:9042...
INFO 22:19:05,267 Using TFramedTransport with a max frame size of 15728640 bytes.
INFO 22:19:05,271 Binding thrift service to /192.168.10.206:9160
INFO 22:19:05,320 Using synchronous/threadpool thrift server on 192.168.10.206 : 9160
INFO 22:19:05,322 Listening for thrift clients...
INFO 22:20:12,302 Handshaking version with /192.168.10.104
INFO 22:20:12,345 Node /192.168.10.104 is now part of the cluster
INFO 22:20:12,422 Handshaking version with /192.168.10.104
INFO 22:20:12,436 InetAddress /192.168.10.104 is now UP

经检查，发现防火墙开着，两节点一直无法连接上。在修改防火墙后，终于成功解决问题。
查看集群状态如下：

[root@localhost bin]# ./nodetool -h 192.168.10.206 ring

Datacenter: datacenter1
==========
Address         Rack        Status State   Load            Owns                Token
                                                                               9219864279989867836
192.168.10.104 rack1       Up     Normal 61.04 KB        100.00%             -9209067550672142833
.............................
192.168.10.104 rack1       Up     Normal 61.04 KB        100.00%             9119365032104954381
192.168.10.104 rack1       Up     Normal 61.04 KB        100.00%             9144131727983125656
192.168.10.104 rack1       Up     Normal 61.04 KB        100.00%             9194205234863680199
192.168.10.104 rack1       Up     Normal 61.04 KB        100.00%             9205150584807094387
192.168.10.206 rack1       Up     Normal 64.28 KB        100.00%             -9093153247508407374
192.168.10.206 rack1       Up     Normal 64.28 KB        100.00%             -8959856259961417723
192.168.10.206 rack1       Up     Normal 64.28 KB        100.00%             -8933270068536399704
........................

Warning: "nodetool ring" is used to output all the tokens of a node.
To view status related info of a node use "nodetool status" instead.

测试：

1.在 206 节点创建 keyspace,并插入数据：

[root@localhost bin]# ./cassandra-cli -host 192.168.10.206
Connected to: "Cluster1" on 192.168.10.206/9160
Welcome to Cassandra CLI version 2.0.6

The CLI is deprecated and will be removed in Cassandra 3.0. Consider migrating to cqlsh.
CQL is fully backwards compatible with Thrift data; see http://www.datastax.com/dev/blog/thrift-to-cql3

Type 'help;' or '?' for help.
Type 'quit;' or 'exit;' to quit.

[default@unknown] create keyspace testspace
... ;
6ceaabf9-afde-3d03-ae18-21da5c18f58b

[default@unknown] use testspace;
Authenticated to keyspace: testspace

[default@testspace] create column family testcf
...      with column_type = 'Standard'
...      and comparator = 'UTF8Type'
...      and default_validation_class = 'UTF8Type'
...      and key_validation_class = 'UTF8Type'
...      and read_repair_chance = 0.1
...      and dclocal_read_repair_chance = 0.0
...      and populate_io_cache_on_flush = false
...      and gc_grace = 864000
...      and min_compaction_threshold = 4
...      and max_compaction_threshold = 32
...      and replicate_on_write = true
...      and compaction_strategy = 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy'
...      and caching = 'KEYS_ONLY'
...      and default_time_to_live = 0
...      and speculative_retry = 'NONE'
...      and compression_options = {'sstable_compression' : 'org.apache.cassandra.io.compress.LZ4Compressor'}
...      and index_interval = 128;
273dc95f-512c-336f-b6e4-fb751560c4f2
[default@testspace] set testcf['num']['num1']=1111;
Value inserted.
Elapsed time: 91 msec(s).
[default@testspace]

2.在节点2查看数据：

[root@localhost bin]# ./cassandra-cli -h 192.168.10.104
Connected to: "Cluster1" on 192.168.10.104/9160
Welcome to Cassandra CLI version 2.0.6

The CLI is deprecated and will be removed in Cassandra 3.0. Consider migrating to cqlsh.
CQL is fully backwards compatible with Thrift data; see http://www.datastax.com/dev/blog/thrift-to-cql3

Type 'help;' or '?' for help.
Type 'quit;' or 'exit;' to quit.

[default@unknown] use testspace;
Authenticated to keyspace: testspace
[default@testspace] list testcf;
Using default limit of 100
Using default cell limit of 100
-------------------
RowKey: num
=> (name=num1, value=1111, timestamp=1397399586622000)

1 Row Returned.
Elapsed time: 421 msec(s).
[default@testspace]

可以看到，数据已同步过来了。