Cassandra集群部署
主机信息
172.28.18.10 mongodb-test7
172.28.18.11 mongodb-test8
172.28.18.20 mongodb-test9
操作系统参数配置
root@test-mysql~]#cat/etc/security/limits.d/90-nproc.conf
*softnproc32000
*hardnproc32000
rootsoftnprocunlimited
*softnofile32000
*hardnofile40960
ulimit的优化
cat/etc/security/limits.conf
*soft nofile 102400
*hard nofile 102400
*soft stack 1024
*hard stack 1024
内核的tcp优化
net.ipv4.tcp_syncookies=1
net.ipv4.tcp_tw_reuse=1
net.ipv4.tcp_tw_recycle=1
net.ipv4.tcp_synack_retries=2
net.ipv4.tcp_syn_retries=2
net.ipv4.tcp_wmem=8192436600873200
net.ipv4.tcp_rmem=32768436600873200
net.ipv4.tcp_mem=945000009150000092700000
net.ipv4.tcp_max_orphans=3276800
net.ipv4.tcp_fin_timeout=30
vm.swappiness=5
#直接生效
/sbin/sysctl-p
python必须安装2.7版本
$tarxvfPython-2.7.11.tgz
mkdir/usr/local/python27
#cdPython-2.7.11
[root@mongodb-test8Python-2.7.11]#./configure--prefix=/usr/local/python27
make
makeinstall
#[python]
exportPY_HOME=/usr/local/python27
exportPATH=$PY_HOME/bin:$PATH
export JAVA_HOME=/usr/local/jdk1.7.0_60
export PATH=$JAVA_HOME/bin:$PATH
exportCLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar
ulimit -S -c unlimited > /dev/null 2>&1
#[cassandra]
export MGO_HOME=/opt/freeware/apache-cassandra-2.2.4
export PATH=$MGO_HOME/bin:$PATH
创建用户目录
#mkdir-p/data/users
添加用户组
groupadd-g355cassandra
添加用户mgousr01
useradd-u367-gcassandra-Gcassandra,users-d/data/users/cassandra01-mcassandra01
修改cassandra用户密码
passwdcassandra01--stdin<<<cassandra01
切换到mgousr01用户
创建如下目录
添加环境变量
#su-cassandra
mkdir -p cassandra/bin
mkdir -p cassandra/dbdata
mkdir -p cassandra/etc
mkdir -p cassandra/logs
mkdir -p cassandra/commitlog
mkdir -p cassandra/saved_caches
添加环境变量
$vim.bash_profile
#[cassandra]
exportMGO_HOME=/opt/freeware/apache-cassandra-2.2.4
exportPATH=$MGO_HOME/bin:$PATH
$source.bash_profile
serviceiptablesstop
chkconfigiptablesoff
#more/etc/hosts
172.28.18.10mongodb-test7
172.28.18.11mongodb-test8
172.28.18.20mongodb-test9
vim /etc/profile
exportJAVA_HOME=/opt/freeware/jdk1.7.0_71
exportPATH=$JAVA_HOME/bin:$PATH
exportCLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar
在mongodb-test7上
编辑配置文件cassandra.yaml
# more cassandra.yaml
# Cassandra storage config YAML
# NOTE:
# See http://wiki.apache.org/cassandra/StorageConfiguration for
# fullexplanations of configuration directives
# /NOTE
# The name of the cluster. This is mainly used to preventmachines in
# one logical cluster from joining another.
cluster_name: 'Cluster01'
# This defines the number of tokens randomly assignedto this node on the ring
# The more tokens, relative to other nodes, the largerthe proportion of data
# that this node will store. You probably want allnodes to have the same number
# of tokens assuming they have equal hardwarecapability.
#
# If you leave this unspecified, Cassandra will usethe default of 1 token for legacy compatibility,
# and will use the initial_token as described below.
#
# Specifying initial_token will override this settingon the node's initial start,
# on subsequent starts, this setting will apply evenif initial token is set.
#
# If you already have a cluster with 1 token per node,and wish to migrate to
# multiple tokens per node, see http://wiki.apache.org/cassandra/Operations
num_tokens: 256
# initial_token allows you to specify tokensmanually. While you can use # it with
# vnodes (num_tokens > 1, above) -- in which caseyou should provide a
# comma-separated list -- it's primarily used whenadding nodes # to legacy clusters
# that do not have vnodes enabled.
# initial_token:
# See http://wiki.apache.org/cassandra/HintedHandoff
# May either be "true" or "false"to enable globally, or contain a list
# of data centers to enable per-datacenter.
# hinted_handoff_enabled: DC1,DC2
hinted_handoff_enabled: true
# this defines the maximum amount of time a dead hostwill have hints
# generated. After it has been dead this long, new hints for it will not be
# created until it has been seen alive and gone downagain.
max_hint_window_in_ms: 10800000 # 3 hours
# Maximum throttle in KBs per second, per deliverythread. This will be
# reduced proportionally to the number of nodes in thecluster. (If there
# are two nodes in the cluster, each delivery threadwill use the maximum
# rate; if there are three, each will throttle to halfof the maximum,
# since we expect two nodes to be delivering hintssimultaneously.)
hinted_handoff_throttle_in_kb: 1024
# Number of threads with which to deliver hints;
# Consider increasing this number when you havemulti-dc deployments, since
# cross-dc handoff tends to be slower
max_hints_delivery_threads: 2
# Maximum throttle in KBs per second, total. This willbe
# reduced proportionally to the number of nodes in thecluster.
batchlog_replay_throttle_in_kb: 1024
# Authentication backend, implementing IAuthenticator;used to identify users
# Out of the box, Cassandra providesorg.apache.cassandra.auth.{AllowAllAuthenticator,
# PasswordAuthenticator}.
#
# - AllowAllAuthenticator performs no checks - set itto disable authentication.
# - PasswordAuthenticator relies on username/passwordpairs to authenticate
# users. Itkeeps usernames and hashed passwords in system_auth.credentials table.
# Pleaseincrease system_auth keyspace replication factor if you use this authenticator.
# If usingPasswordAuthenticator, CassandraRoleManager must also be used (see below)
authenticator: AllowAllAuthenticator
# Authorization backend, implementing IAuthorizer;used to limit access/provide permissions
# Out of the box, Cassandra providesorg.apache.cassandra.auth.{AllowAllAuthorizer,
# CassandraAuthorizer}.
#
# - AllowAllAuthorizer allows any action to any user -set it to disable authorization.
# - CassandraAuthorizer stores permissions insystem_auth.permissions table. Please
# increasesystem_auth keyspace replication factor if you use this authorizer.
authorizer: AllowAllAuthorizer
# Part of the Authentication & Authorizationbackend, implementing IRoleManager; used
# to maintain grants and memberships between roles.
# Out of the box, Cassandra providesorg.apache.cassandra.auth.CassandraRoleManager,
# which stores role information in the system_authkeyspace. Most functions of the
# IRoleManager require an authenticated login, sounless the configured IAuthenticator
# actually implements authentication, most of thisfunctionality will be unavailable.
#
# - CassandraRoleManager stores role data in thesystem_auth keyspace. Please
# increasesystem_auth keyspace replication factor if you use this role manager.
role_manager: CassandraRoleManager
# Validity period for roles cache (fetchingpermissions can be an
# expensive operation depending on the authorizer).Granted roles are cached for
# authenticated sessions in AuthenticatedUser andafter the period specified
# here, become eligible for (async) reload.
# Defaults to 2000, set to 0 to disable.
# Will be disabled automatically forAllowAllAuthenticator.
roles_validity_in_ms: 2000
# Refresh interval for roles cache (if enabled).
# After this interval, cache entries become eligiblefor refresh. Upon next
# access, an async reload is scheduled and the oldvalue returned until it
# completes. If roles_validity_in_ms is non-zero, thenthis must be
# also.
# Defaults to the same value as roles_validity_in_ms.
# roles_update_interval_in_ms: 1000
# Validity period for permissions cache (fetchingpermissions can be an
# expensive operation depending on the authorizer,CassandraAuthorizer is
# one example). Defaults to 2000, set to 0 to disable.
# Will be disabled automatically forAllowAllAuthorizer.
permissions_validity_in_ms: 2000
# Refresh interval for permissions cache (if enabled).
# After this interval, cache entries become eligiblefor refresh. Upon next
# access, an async reload is scheduled and the oldvalue returned until it
# completes. If permissions_validity_in_ms isnon-zero, then this must be
# also.
# Defaults to the same value aspermissions_validity_in_ms.
# permissions_update_interval_in_ms: 1000
# The partitioner is responsible for distributinggroups of rows (by
# partition key) across nodes in the cluster. You should leave this
# alone for new clusters. The partitioner can NOT be changed without
# reloading all data, so when upgrading you should setthis to the
# same partitioner you were already using.
#
# Besides Murmur3Partitioner, partitioners includedfor backwards
# compatibility include RandomPartitioner,ByteOrderedPartitioner, and
# OrderPreservingPartitioner.
#
partitioner: org.apache.cassandra.dht.Murmur3Partitioner
# Directories where Cassandra should store data ondisk. Cassandra
# will spread data evenly across them, subject to thegranularity of
# the configured compaction strategy.
# If not set, the default directory is$CASSANDRA_HOME/data/data.
data_file_directories:
# - /var/lib/cassandra/data
-/data/users/cassandra01/cassandra/dbdata
# commit log. when running on magnetic HDD, this should be a
# separate spindle than the data directories.
# If not set, the default directory is$CASSANDRA_HOME/data/commitlog.
# commitlog_directory: /var/lib/cassandra/commitlog
# ommitlog_directory:/data/users/cassandra01/cassandra/commitlog
# 在10节点要注释(不注释启动会报错)，在11和20节点不要注释
# policy for data disk failures:
# die: shut down gossip and client transports and killthe JVM for any fs errors or
# single-sstable errors, so the node can be replaced.
# stop_paranoid: shut down gossip and clienttransports even for single-sstable errors,
# kill the JVM for errors during startup.
# stop: shut down gossip and client transports,leaving the node effectively dead, but
# canstill be inspected via JMX, kill the JVM for errors during startup.
# best_effort: stop using the failed disk and respondto requests based on
# remaining available sstables. This means you WILL see obsolete
# data at CL.ONE!
# ignore: ignore fatal errors and let requests fail,as in pre-1.2 Cassandra
disk_failure_policy: stop
# policy for commit disk failures:
# die: shut down gossip and Thrift and kill the JVM,so the node can be replaced.
# stop: shut down gossip and Thrift, leaving the nodeeffectively dead, but
# canstill be inspected via JMX.
# stop_commit: shutdown the commit log, letting writescollect but
# continuing to service reads, as in pre-2.0.5 Cassandra
# ignore: ignore fatal errors and let the batches fail
commit_failure_policy: stop
# Maximum size of the key cache in memory.
#
# Each key cache hit saves 1 seek and each row cachehit saves 2 seeks at the
# minimum, sometimes more. The key cache is fairlytiny for the amount of
# time it saves, so it's worthwhile to use it at largenumbers.
# The row cache saves even more time, but must containthe entire row,
# so it is extremely space-intensive. It's best toonly use the
# row cache if you have hot rows or static rows.
#
# NOTE: if you reduce the size, you may not get youhottest keys loaded on startup.
#
# Default value is empty to make it "auto"(min(5% of Heap (in MB), 100MB)). Set to 0 to disable key ca
che.
key_cache_size_in_mb:
# Duration in seconds after which Cassandra should
# save the key cache. Caches are saved tosaved_caches_directory as
# specified in this configuration file.
#
# Saved caches greatly improve cold-start speeds, andis relatively cheap in
# terms of I/O for the key cache. Row cache saving ismuch more expensive and
# has limited use.
#
# Default is 14400 or 4 hours.
key_cache_save_period: 14400
# Number of keys from the key cache to save
# Disabled by default, meaning all keys are going tobe saved
# key_cache_keys_to_save: 100
# Row cache implementation class name.
# Available implementations:
# org.apache.cassandra.cache.OHCProvider Fully off-heap row cacheimplementation (defa
ult).
# org.apache.cassandra.cache.SerializingCacheProvider This is the row cache implementationavailabi
le
# in previous releases of Cassandra.
# row_cache_class_name:org.apache.cassandra.cache.OHCProvider
# Maximum size of the row cache in memory.
# Please note that OHC cache implementation requiressome additional off-heap memory to manage
# the map structures and some in-flight memory duringoperations before/after cache entries can be
# accounted against the cache capacity. This overheadis usually small compared to the whole capacity.
# Do not specify more memory that the system canafford in the worst usual situation and leave some
# headroom for OS block level cache. Do never allowyour system to swap.
#
# Default value is 0, to disable row caching.
row_cache_size_in_mb: 0
# Duration in seconds after which Cassandra shouldsave the row cache.
# Caches are saved to saved_caches_directory asspecified in this configuration file.
#
# Saved caches greatly improve cold-start speeds, andis relatively cheap in
# terms of I/O for the key cache. Row cache saving ismuch more expensive and
# has limited use.
#
# Default is 0 to disable saving the row cache.
row_cache_save_period: 0
# Number of keys from the row cache to save.
# Specify 0 (which is the default), meaning all keysare going to be saved
# row_cache_keys_to_save: 100
# Maximum size of the counter cache in memory.
#
# Counter cache helps to reduce counter locks'contention for hot counter cells.
# In case of RF = 1 a counter cache hit will causeCassandra to skip the read before
# write entirely. With RF > 1 a counter cache hitwill still help to reduce the duration
# of the lock hold, helping with hot counter cellupdates, but will not allow skipping
# the read entirely. Only the local (clock, count)tuple of a counter cell is kept
# in memory, not the whole counter, so it's relativelycheap.
#
# NOTE: if you reduce the size, you may not get youhottest keys loaded on startup.
#
# Default value is empty to make it "auto"(min(2.5% of Heap (in MB), 50MB)). Set to 0 to disable count
er cache.
# NOTE: if you perform counter deletes and rely on lowgcgs, you should disable the counter cache.
counter_cache_size_in_mb:
# Duration in seconds after which Cassandra should
# save the counter cache (keys only). Caches are savedto saved_caches_directory as
# specified in this configuration file.
#
# Default is 7200 or 2 hours.
counter_cache_save_period: 7200
# Number of keys from the counter cache to save
# Disabled by default, meaning all keys are going tobe saved
# counter_cache_keys_to_save: 100
# saved caches
# If not set, the default directory is$CASSANDRA_HOME/data/saved_caches.
# saved_caches_directory:/var/lib/cassandra/saved_caches
saved_caches_directory:/data/users/cassandra01/cassandra/saved_caches
# commitlog_sync may be either "periodic" or"batch."
#
# When in batch mode, Cassandra won't ack writes untilthe commit log
# has been fsynced to disk. It will wait
# commitlog_sync_batch_window_in_ms millisecondsbetween fsyncs.
# This window should be kept short because the writerthreads will
# be unable to do extra work while waiting. (You may need to increase
# concurrent_writes for the same reason.)
#
# commitlog_sync: batch
# commitlog_sync_batch_window_in_ms: 2
#
# the other option is "periodic" wherewrites may be acked immediately
# and the CommitLog is simply synced everycommitlog_sync_period_in_ms
# milliseconds.
commitlog_sync: periodic
commitlog_sync_period_in_ms: 10000
# The size of the individual commitlog filesegments. A commitlog
# segment may be archived, deleted, or recycled onceall the data
# in it (potentially from each columnfamily in thesystem) has been
# flushed to sstables.
#
# The default size is 32, which is almost always fine,but if you are
# archiving commitlog segments (seecommitlog_archiving.properties),
# then you probably want a finer granularity ofarchiving; 8 or 16 MB
# is reasonable.
commitlog_segment_size_in_mb: 32
# Compression to apply to the commit log. If omitted,the commit log
# will be written uncompressed. LZ4, Snappy, and Deflate compressors
# are supported.
#commitlog_compression:
# -class_name: LZ4Compressor
# parameters:
# -
# any class that implements the SeedProvider interfaceand has a
# constructor that takes a Map<String, String>of parameters will do.
seed_provider:
# Addressesof hosts that are deemed contact points.
# Cassandra nodes use this list of hosts tofind each other and learn
# thetopology of the ring. You must changethis if you are running
# multiplenodes!
-class_name: org.apache.cassandra.locator.SimpleSeedProvider
parameters:
#seeds is actually a comma-delimited list of addresses.
# Ex:"<ip1>,<ip2>,<ip3>"
- seeds: "mongodb-test8,mongodb-test9" #11和20节点主机名或IP
# For workloads with more data than can fit in memory,Cassandra's
# bottleneck will be reads that need to fetch datafrom
# disk. "concurrent_reads" should be set to(16 * number_of_drives) in
# order to allow the operations to enqueue low enoughin the stack
# that the OS and drives can reorder them. Sameapplies to
# "concurrent_counter_writes", since counterwrites read the current
# values before incrementing and writing them back.
#
# On the other hand, since writes are almost never IObound, the ideal
# number of "concurrent_writes" is dependenton the number of cores in
# your system; (8 * number_of_cores) is a good rule ofthumb.
concurrent_reads: 32
concurrent_writes: 32
concurrent_counter_writes: 32
# Total memory to use for sstable-readingbuffers. Defaults to
# the smaller of 1/4 of heap or 512MB.
# file_cache_size_in_mb: 512
# Total permitted memory to use for memtables.Cassandra will stop
# accepting writes when the limit is exceeded until aflush completes,
# and will trigger a flush based onmemtable_cleanup_threshold
# If omitted, Cassandra will set both to 1/4 the sizeof the heap.
# memtable_heap_space_in_mb: 2048
# memtable_offheap_space_in_mb: 2048
# Ratio of occupied non-flushing memtable size tototal permitted size
# that will trigger a flush of the largest memtable.Larger mct will
# mean larger flushes and hence less compaction, butalso less concurrent
# flush activity which can make it difficult to keepyour disks fed
# under heavy write load.
#
# memtable_cleanup_threshold defaults to 1 /(memtable_flush_writers + 1)
# memtable_cleanup_threshold: 0.11
# Specify the way Cassandra allocates and managesmemtable memory.
# Options are:
# heap_buffers: on heap niobuffers
# offheap_buffers: off heap (direct) nio buffers
# offheap_objects: native memory, eliminating nio buffer heap overhead
memtable_allocation_type: heap_buffers
# Total space to use for commit logs on disk.
#
# If space gets above this value, Cassandra will flushevery dirty CF
# in the oldest segment and remove it. So a small total commitlog space
# will tend to cause more flush activity onless-active columnfamilies.
#
# The default value is the smaller of 8192, and 1/4 ofthe total space
# of the commitlog volume.
#
# commitlog_total_space_in_mb: 8192
# This sets the amount of memtable flush writerthreads. These will
# be blocked by disk io, and each one will hold amemtable in memory
# while blocked.
#
# memtable_flush_writers defaults to the smaller of(number of disks,
# number of cores), with a minimum of 2 and a maximumof 8.
#
# If your data directories are backed by SSD, youshould increase this
# to the number of cores.
#memtable_flush_writers: 8
# A fixed memory pool size in MB for for SSTable indexsummaries. If left
# empty, this will default to 5% of the heap size. Ifthe memory usage of
# all index summaries exceeds this limit, SSTableswith low read rates will
# shrink their index summaries in order to meet thislimit. However, this
# is a best-effort process. In extreme conditionsCassandra may need to use
# more than this amount of memory.
index_summary_capacity_in_mb:
# How frequently index summaries should beresampled. This is done
# periodically to redistribute memory from thefixed-size pool to sstables
# proportional their recent read rates. Setting to -1 will disable this
# process, leaving existing index summaries at theircurrent sampling level.
index_summary_resize_interval_in_minutes: 60
# Whether to, when doing sequential writing, fsync()at intervals in
# order to force the operating system to flush thedirty
# buffers. Enable this to avoid sudden dirty bufferflushing from
# impacting read latencies. Almost always a good ideaon SSDs; not
# necessarily on platters.
trickle_fsync: false
trickle_fsync_interval_in_kb: 10240
# TCP port, for commands and data
# For security reasons, you should not expose thisport to the internet. Firewall it ifneeded.
storage_port: 7000
# SSL port, for encrypted communication. Unused unless enabled in
# encryption_options
# For security reasons, you should not expose thisport to the internet. Firewall it ifneeded.
ssl_storage_port: 7001
# Address or interface to bind to and tell otherCassandra nodes to connect to.
# You _must_ change this if you want multiple nodes tobe able to communicate!
#
# Set listen_address OR listen_interface, not both.Interfaces must correspond
# to a single address, IP aliasing is not supported.
#
# Leaving it blank leaves it up toInetAddress.getLocalHost(). This
# will always do the Right Thing _if_ the node isproperly configured
# (hostname, name resolution, etc), and the RightThing is to use the
# address associated with the hostname (it might notbe).
#
# Setting listen_address to 0.0.0.0 is always wrong.
#
# If you choose to specify the interface by name andthe interface has an ipv4 and an ipv6 address
# you can specify which should be chosen usinglisten_interface_prefer_ipv6. If false the first ipv4
# address will be used. If true the first ipv6 addresswill be used. Defaults to false preferring
# ipv4. If there is only one address it will beselected regardless of ipv4/ipv6.
listen_address: 172.28.18.10
# listen_interface: eth0
# listen_interface_prefer_ipv6: false
# Address to broadcast to other Cassandra nodes
# Leaving this blank will set it to the same value aslisten_address
# broadcast_address: 1.2.3.4
# Internode authentication backend, implementingIInternodeAuthenticator;
# used to allow/disallow connections from peer nodes.
# internode_authenticator:org.apache.cassandra.auth.AllowAllInternodeAuthenticator
# Whether to start the native transport server.
# Please note that the address on which the nativetransport is bound is the
# same as the rpc_address. The port however isdifferent and specified below.
start_native_transport: true
# port for the CQL native transport to listen forclients on
# For security reasons, you should not expose thisport to the internet. Firewall it ifneeded.
native_transport_port: 9042
# The maximum threads for handling requests when thenative transport is used.
# This is similar to rpc_max_threads though thedefault differs slightly (and
# there is no native_transport_min_threads, idlethreads will always be stopped
# after 30 seconds).
# native_transport_max_threads: 128
#
# The maximum size of allowed frame. Frame (requests)larger than this will
# be rejected as invalid. The default is 256MB.
# native_transport_max_frame_size_in_mb: 256
# The maximum number of concurrent client connections.
# The default is -1, which means unlimited.
# native_transport_max_concurrent_connections: -1
# The maximum number of concurrent client connectionsper source ip.
# The default is -1, which means unlimited.
# native_transport_max_concurrent_connections_per_ip:-1
# Whether to start the thrift rpc server.
start_rpc: false
# The address or interface to bind the Thrift RPCservice and native transport
# server to.
#
# Set rpc_address OR rpc_interface, not both.Interfaces must correspond
# to a single address, IP aliasing is not supported.
#
# Leaving rpc_address blank has the same effect as onlisten_address
# (i.e. it will be based on the configured hostname ofthe node).
#
# Note that unlike listen_address, you can specify0.0.0.0, but you must also
# set broadcast_rpc_address to a value other than0.0.0.0.
#
# For security reasons, you should not expose thisport to the internet. Firewall it ifneeded.
#
# If you choose to specify the interface by name andthe interface has an ipv4 and an ipv6 address
# you can specify which should be chosen usingrpc_interface_prefer_ipv6. If false the first ipv4
# address will be used. If true the first ipv6 addresswill be used. Defaults to false preferring
# ipv4. If there is only one address it will beselected regardless of ipv4/ipv6.
rpc_address: 172.28.18.10
# rpc_interface: eth1
# rpc_interface_prefer_ipv6: false
# port for Thrift to listen for clients on
rpc_port: 9160
# RPC address to broadcast to drivers and otherCassandra nodes. This cannot
# be set to 0.0.0.0. If left blank, this will be setto the value of
# rpc_address. If rpc_address is set to 0.0.0.0,broadcast_rpc_address must
# be set.
# broadcast_rpc_address: 1.2.3.4
# enable or disable keepalive on rpc/nativeconnections
rpc_keepalive: true
# Cassandra provides two out-of-the-box options forthe RPC Server:
#
# sync ->One thread per thrift connection. For a very large number of clients, memory
# willbe your limiting factor. On a 64 bit JVM, 180KB is the minimum stack size
# per thread, and that will correspond to youruse of virtual memory (but physical memory
# maybe limited depending on use of stack space).
#
# hsha ->Stands for "half synchronous, half asynchronous." All thrift clientsare handled
# asynchronouslyusing a small number of threads that does not vary with the amount
# ofthrift clients (and thus scales well to many clients). The rpc requests arestill
# synchronous (one thread per active request). If hsha is selected then itis essential
# thatrpc_max_threads is changed from the default value of unlimited.
#
# The default is sync because on Windows hsha is about30% slower. On Linux,
# sync/hsha performance is about the same, with hshaof course using less memory.
#
# Alternatively, can provide your own RPC server by providing the fully-qualified classname
# of an o.a.c.t.TServerFactory that can create aninstance of it.
rpc_server_type: sync
# Uncomment rpc_min|max_thread to set request poolsize limits.
#
# Regardless of your choice of RPC server (see above),the number of maximum requests in the
# RPC thread pool dictates how many concurrentrequests are possible (but if you are using the sync
# RPC server, it also dictates the number of clientsthat can be connected at all).
#
# The default is unlimited and thus provides noprotection against clients overwhelming the server. You
are
# encouraged to set a maximum that makes sense for youin production, but do keep in mind that
# rpc_max_threads represents the maximum number ofclient requests this server may execute concurrently
.
#
# rpc_min_threads: 16
# rpc_max_threads: 2048
# uncomment to set socket buffer sizes on rpcconnections
# rpc_send_buff_size_in_bytes:
# rpc_recv_buff_size_in_bytes:
# Uncomment to set socket buffer size for internodecommunication
# Note that when setting this, the buffer size islimited by net.core.wmem_max
# and when not setting it it is defined bynet.ipv4.tcp_wmem
# See:
# /proc/sys/net/core/wmem_max
# /proc/sys/net/core/rmem_max
# /proc/sys/net/ipv4/tcp_wmem
# /proc/sys/net/ipv4/tcp_wmem
# and: man tcp
# internode_send_buff_size_in_bytes:
# internode_recv_buff_size_in_bytes:
# Frame size for thrift (maximum message length).
thrift_framed_transport_size_in_mb: 15
# Set to true to have Cassandra create a hard link toeach sstable
# flushed or streamed locally in a backups/subdirectory of the
# keyspace data. Removing these links is the operator's
# responsibility.
incremental_backups: false
# Whether or not to take a snapshot before eachcompaction. Be
# careful using this option, since Cassandra won'tclean up the
# snapshots for you. Mostly useful if you're paranoid when there
# is a data format change.
snapshot_before_compaction: false
# Whether or not a snapshot is taken of the databefore keyspace truncation
# or dropping of column families. The STRONGLY adviseddefault of true
# should be used to provide data safety. If you setthis flag to false, you will
# lose data on truncation or drop.
auto_snapshot: true
# When executing a scan, within or across a partition,we need to keep the
# tombstones seen in memory so we can return them tothe coordinator, which
# will use them to make sure other replicas also knowabout the deleted rows.
# With workloads that generate a lot of tombstones,this can cause performance
# problems and even exaust the server heap.
# (http://www.datastax.com/dev/blog/cassandra-anti-patterns-queues-and-queue-like-datasets)
# Adjust the thresholds here if you understand thedangers and want to
# scan more tombstones anyway. These thresholds may also be adjusted atruntime
# using the StorageService mbean.
tombstone_warn_threshold: 1000
tombstone_failure_threshold: 100000
# Granularity of the collation index of rows within apartition.
# Increase if your rows are large, or if you have avery large
# number of rows per partition. The competing goals are these:
# 1) a smallergranularity means more index entries are generated
# andlooking up rows withing the partition by collation column
# is faster
# 2) but,Cassandra will keep the collation index in memory for hot
# rows (aspart of the key cache), so a larger granularity means
# you cancache more hot rows
column_index_size_in_kb: 64
# Log WARN on any batch size exceeding this value. 5kbper batch by default.
# Caution should be taken on increasing the size ofthis threshold as it can lead to node instability.
batch_size_warn_threshold_in_kb: 5
# Fail any batch exceeding this value. 50kb (10x warnthreshold) by default.
batch_size_fail_threshold_in_kb: 50
# Number of simultaneous compactions to allow, NOTincluding
# validation "compactions" for anti-entropyrepair. Simultaneous
# compactions can help preserve read performance in amixed read/write
# workload, by mitigating the tendency of smallsstables to accumulate
# during a single long running compactions. Thedefault is usually
# fine and if you experience problems with compactionrunning too
# slowly or too fast, you should look at
# compaction_throughput_mb_per_sec first.
#
# concurrent_compactors defaults to the smaller of(number of disks,
# number of cores), with a minimum of 2 and a maximumof 8.
#
# If your data directories are backed by SSD, youshould increase this
# to the number of cores.
#concurrent_compactors: 1
# Throttles compaction to the given total throughputacross the entire
# system. The faster you insert data, the faster youneed to compact in
# order to keep the sstable count down, but ingeneral, setting this to
# 16 to 32 times the rate you are inserting data ismore than sufficient.
# Setting this to 0 disables throttling. Note thatthis account for all types
# of compaction, including validation compaction.
compaction_throughput_mb_per_sec: 16
# Log a warning when compacting partitions larger thanthis value
compaction_large_partition_warning_threshold_mb: 100
# When compacting, the replacement sstable(s) can beopened before they
# are completely written, and used in place of theprior sstables for
# any range that has been written. This helps tosmoothly transfer reads
# between the sstables, reducing page cache churn andkeeping hot rows hot
sstable_preemptive_open_interval_in_mb: 50
# Throttles all outbound streaming file transfers onthis node to the
# given total throughput in Mbps. This is necessarybecause Cassandra does
# mostly sequential IO when streaming data duringbootstrap or repair, which
# can lead to saturating the network connection anddegrading rpc performance.
# When unset, the default is 200 Mbps or 25 MB/s.
# stream_throughput_outbound_megabits_per_sec: 200
# Throttles all streaming file transfer between thedatacenters,
# this setting allows users to throttle inter dcstream throughput in addition
# to throttling all network stream traffic asconfigured with
# stream_throughput_outbound_megabits_per_sec
#inter_dc_stream_throughput_outbound_megabits_per_sec:
# How long the coordinator should wait for readoperations to complete
read_request_timeout_in_ms: 5000
# How long the coordinator should wait for seq orindex scans to complete
range_request_timeout_in_ms: 10000
# How long the coordinator should wait for writes tocomplete
write_request_timeout_in_ms: 2000
# How long the coordinator should wait for counterwrites to complete
counter_write_request_timeout_in_ms: 5000
# How long a coordinator should continue to retry aCAS operation
# that contends with other proposals for the same row
cas_contention_timeout_in_ms: 1000
# How long the coordinator should wait for truncatesto complete
# (This can be much longer, because unlessauto_snapshot is disabled
# we need to flush first so we can snapshot beforeremoving the data.)
truncate_request_timeout_in_ms: 60000
# The default timeout for other, miscellaneousoperations
request_timeout_in_ms: 10000
# Enable operation timeout information exchangebetween nodes to accurately
# measure request timeouts. If disabled, replicas will assume thatrequests
# were forwarded to them instantly by the coordinator,which means that
# under overload conditions we will waste that muchextra time processing
# already-timed-out requests.
#
# Warning: before enabling this property make sure tontp is installed
# and the times are synchronized between the nodes.
cross_node_timeout: false
# Enable socket timeout for streaming operation.
# When a timeout occurs during streaming, streaming isretried from the start
# of the current file. This _can_ involve re-streamingan important amount of
# data, so you should avoid setting the value too low.
# Default value is 3600000, which means streamstimeout after an hour.
# streaming_socket_timeout_in_ms: 3600000
# phi value that must be reached for a host to bemarked down.
# most users should never need to adjust this.
# phi_convict_threshold: 8
# endpoint_snitch -- Set this to a class thatimplements
# IEndpointSnitch. The snitch has two functions:
# - it teaches Cassandra enough about your networktopology to route
# requestsefficiently
# - it allows Cassandra to spread replicas around yourcluster to avoid
# correlatedfailures. It does this by grouping machines into
# "datacenters" and "racks." Cassandra will do its best not to have
# more thanone replica on the same "rack" (which may not actually
# be aphysical location)
#
# IF YOU CHANGE THE SNITCH AFTER DATA IS INSERTED INTOTHE CLUSTER,
# YOU MUST RUN A FULL REPAIR, SINCE THE SNITCH AFFECTSWHERE REPLICAS
# ARE PLACED.
#
# IF THE RACK A REPLICA IS PLACED IN CHANGES AFTER THEREPLICA HAS BEEN
# ADDED TO A RING, THE NODE MUST BE DECOMMISSIONED ANDREBOOTSTRAPPED.
#
# Out of the box, Cassandra provides
# -SimpleSnitch:
# TreatsStrategy order as proximity. This can improve cache
# localitywhen disabling read repair. Onlyappropriate for
# single-datacenter deployments.
# -GossipingPropertyFileSnitch
# This shouldbe your go-to snitch for production use. The rack
# anddatacenter for the local node are defined in
# cassandra-rackdc.properties and propagated to other nodes via
# gossip. Ifcassandra-topology.properties exists, it is used as a
# fallback,allowing migration from the PropertyFileSnitch.
# -PropertyFileSnitch:
# Proximityis determined by rack and data center, which are
# explicitlyconfigured in cassandra-topology.properties.
# - Ec2Snitch:
# Appropriatefor EC2 deployments in a single Region. Loads Region
# andAvailability Zone information from the EC2 API. The Region is
# treated asthe datacenter, and the Availability Zone as the rack.
# Onlyprivate IPs are used, so this will not work across multiple
# Regions.
# -Ec2MultiRegionSnitch:
# Uses publicIPs as broadcast_address to allow cross-region
# connectivity. (Thus, you shouldset seed addresses to the public
# IP aswell.) You will need to open the storage_port or
# ssl_storage_port on the public IP firewall. (For intra-Region
# traffic,Cassandra will switch to the private IP after
# establishing a connection.)
# -RackInferringSnitch:
# Proximityis determined by rack and data center, which are
# assumed tocorrespond to the 3rd and 2nd octet of each node's IP
# address,respectively. Unless this happens tomatch your
# deploymentconventions, this is best used as an example of
# writing acustom Snitch class and is provided in that spirit.
#
# You can use a custom Snitch by setting this to thefull class name
# of the snitch, which will be assumed to be on yourclasspath.
endpoint_snitch: SimpleSnitch #假设所有服务器都在同一个机架里面，就没有机架策略了；假设所有机器都是平等的，分布在同一个局域网里面。
# controls how often to perform the more expensivepart of host score
# calculation
dynamic_snitch_update_interval_in_ms: 100
# controls how often to reset all host scores,allowing a bad host to
# possibly recover
dynamic_snitch_reset_interval_in_ms: 600000
# if set greater than zero and read_repair_chance is< 1.0, this will allow
# 'pinning' of replicas to hosts in order to increasecache capacity.
# The badness threshold will control how much worsethe pinned host has to be
# before the dynamic snitch will prefer other replicasover it. This is
# expressed as a double which represents apercentage. Thus, a value of
# 0.2 means Cassandra would continue to prefer thestatic snitch values
# until the pinned host was 20% worse than thefastest.
dynamic_snitch_badness_threshold: 0.1
# request_scheduler -- Set this to a class thatimplements
# RequestScheduler, which will schedule incoming clientrequests
# according to the specific policy. This is useful formulti-tenancy
# with a single Cassandra cluster.
# NOTE: This is specifically for requests from theclient and does
# not affect inter node communication.
# org.apache.cassandra.scheduler.NoScheduler - Noscheduling takes place
# org.apache.cassandra.scheduler.RoundRobinScheduler -Round robin of
# client requests to a node with a separate queue foreach
# request_scheduler_id. The scheduler is furthercustomized by
# request_scheduler_options as described below.
request_scheduler:org.apache.cassandra.scheduler.NoScheduler
# Scheduler Options vary based on the type ofscheduler
# NoScheduler - Has no options
# RoundRobin
# -throttle_limit -- The throttle_limit is the number of in-flight
# requests per client. Requestsbeyond
# that limit are queued up until
# running requests can complete.
# The value of 80 here is twice the number of
# concurrent_reads + concurrent_writes.
# -default_weight -- default_weight is optional and allows for
# overriding the default which is 1.
# - weights --Weights are optional and will default to 1 or the
# overridden default_weight. The weight translates into how
# many requests are handled during each turn of the
# RoundRobin, based on the scheduler id.
#
# request_scheduler_options:
# throttle_limit: 80
# default_weight: 5
# weights:
# Keyspace1: 1
# Keyspace2: 5
# request_scheduler_id -- An identifier based on whichto perform
# the request scheduling. Currently the only validoption is keyspace.
# request_scheduler_id: keyspace
# Enable or disable inter-node encryption
# Default settings are TLS v1, RSA 1024-bit keys (itis imperative that
# users generate their own keys)TLS_RSA_WITH_AES_128_CBC_SHA as the cipher
# suite for authentication, key exchange andencryption of the actual data transfers.
# Use the DHE/ECDHE ciphers if running in FIPS 140compliant mode.
# NOTE: No custom encryption options are enabled atthe moment
# The available internode options are : all, none, dc,rack
#
# If set to dc cassandra will encrypt the trafficbetween the DCs
# If set to rack cassandra will encrypt the trafficbetween the racks
#
# The passwords used in these options must match thepasswords used when generating
# the keystore and truststore. For instructions on generating these files,see:
ore
#
server_encryption_options:
internode_encryption:none
keystore:conf/.keystore
keystore_password: cassandra
truststore:conf/.truststore
truststore_password: cassandra
# Moreadvanced defaults below:
# protocol:TLS
# algorithm:SunX509
#store_type: JKS
#cipher_suites:[TLS_RSA_WITH_AES_128_CBC_SHA,TLS_RSA_WITH_AES_256_CBC_SHA,TLS_DHE_RSA_WITH_AES_12
8_CBC_SHA,TLS_DHE_RSA_WITH_AES_256_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_25
6_CBC_SHA]
#require_client_auth: false
# enable or disable client/server encryption.
client_encryption_options:
enabled:false
# If enabledand optional is set to true encrypted and unencrypted connections are handled.
optional:false
keystore:conf/.keystore
keystore_password: cassandra
#require_client_auth: false
# Settrustore and truststore_password if require_client_auth is true
#truststore: conf/.truststore
#truststore_password: cassandra
# Moreadvanced defaults below:
# protocol:TLS
# algorithm:SunX509
#store_type: JKS
#cipher_suites:[TLS_RSA_WITH_AES_128_CBC_SHA,TLS_RSA_WITH_AES_256_CBC_SHA,TLS_DHE_RSA_WITH_AES_12
8_CBC_SHA,TLS_DHE_RSA_WITH_AES_256_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA,TLS_ECDHE_RSA_WITH_AES_25
6_CBC_SHA]
# internode_compression controls whether trafficbetween nodes is
# compressed.
# can be: all - all traffic is compressed
# dc - traffic between differentdatacenters is compressed
# none- nothing is compressed.
internode_compression: all
# Enable or disable tcp_nodelay for inter-dccommunication.
# Disabling it will result in larger (but fewer)network packets being sent,
# reducing overhead from the TCP protocol itself, atthe cost of increasing
# latency if you block for cross-datacenter responses.
inter_dc_tcp_nodelay: false
# TTL for different trace types used during logging ofthe repair process.
tracetype_query_ttl: 86400
tracetype_repair_ttl: 604800
# GC Pauses greater than gc_warn_threshold_in_ms willbe logged at WARN level
# Adjust the threshold based on your applicationthroughput requirement
# By default, Cassandra logs GC Pauses greater than200 ms at INFO level
# gc_warn_threshold_in_ms: 1000
# UDFs (user defined functions) are disabled bydefault.
# As of Cassandra 2.2, there is no security manager oranything else in place that
# prevents execution of evil code. CASSANDRA-9402 willfix this issue for Cassandra 3.0.
# This will inherently be backwards-incompatible withany 2.2 UDF that perform insecure
# operations such as opening a socket or writing tothe filesystem.
enable_user_defined_functions: false
# The default Windows kernel timer and schedulingresolution is 15.6ms for power conservation.
# Lowering this value on Windows can provide much tighterlatency and better throughput, however
# some virtualized environments may see a negativeperformance impact from changing this setting
# below their system default. The sysinternals'clockres' tool can confirm your system's default
# setting.
windows_timer_interval: 1
、、、、、、
endpoint_snitch:SimpleSnitch--假设所有服务器都在同一个机架里面，就没有机架策略了；
假设所有机器都是平等的，分布在同一个局域网里面。
编辑配置文件logback.xml
<file>/data/users/cassandra01/cassandra/logs/system.log</file>
配置cassandra-env.sh
vicassandra-env.sh
MAX_HEAP_SIZE根据系统物理内存设置;HEAP_NEWSIZE设置为MAX_HEAP_SIZE/4;这两个参数会自动设置,可以去掉注释指定具体值.
----------------------------------------------------------
MAX_HEAP_SIZE="4G"
HEAP_NEWSIZE="800M"
配置完毕10节点，将配置好的scp到11和20节点
$scp-rcassandra/cassandra01@mongodb-test8:/data/users/cassandra01
$scp-rcassandra/cassandra01@mongodb-test9:/data/users/cassandra01
在各个节点上对应修改cassandra.yaml(只修改如下)
11主机
listen_address:172.28.18.11
rpc_address:172.28.18.11
20主机
listen_address:172.28.18.20
rpc_address:172.28.18.20
按顺序启动
先seeds，后main(10主机)
在11主机
cassandra-f
在20主机
cassandra-f
在10主机
cassandra-f
这些启动都是在前台启动，使用&在后台启动运行
在10主机main启动cli
(需要事先安装python2.7以上版本)
cqlsh mongodb-test7
测试
create keyspace mykeyspacewith replication={'class':'SimpleStrategy','replication_factor':1};
【
create keyspace excelsior with replication ={'calss':'SimpleStrategy','replication_factor':3};
create keyspace "Excalibur" with replication= {'class':'NetworkTopologyStrategy','dc1':3,'dc2':2};
第一个副本因子是3个，第二个副本因子是2个；
】
use mykeyspace;
create table users(userid int primary key, fn text, lntext, number int);
insert into users(userid, number) values(1 ,2);
insert into users(userid, fn, ln, number) values(1, 'a','b', 3);
select avg(number) from users;
注意：
如果出现这样的错误：
No single argument constructor found for class[Ljava.lang.String;……
有可能是cassandra.yaml配置文件的格式有问题
注意！
data_file_directories:
- /cassandradata/data
insert into users(userid, fn, ln, number) values(2,'2a', '2b', 23);
insert into users(userid, fn, ln, number) values(3,'2a', '2b', 33);
insert into users(userid, fn, ln, number) values(4,'2a', '2b', 33);
insert into users(userid, fn, ln, number) values(5,'2a', '2b', 33);
insert into users(userid, fn, ln, number) values(6,'2a', '2b', 33);
insert into users(userid, fn, ln, number) values(7,'2a', '2b', 33);
insert into users(userid, fn, ln, number) values(8,'2a', '2b', 33);
本文出自 “九指神丐” 博客，请务必保留此出处http://beigai.blog.51cto.com/8308160/1749531