Elasticsearch启动分析与问题解决-bootstrap checks

0 说明

使用的es版本为5.6，Linux版本为CentOs 6.5.

1 Elasticsearch bootstrap checks

1.1 开发环境

如果在es的配置中没有配置network.host来指定一个可用的IP地址的话，默认情况下，就绑定到localhost上，此时es会认为用户只是在开发环境下使用es，基于开箱即用的原则，虽然es此时也会进行bootstrap checks，来检查用户的配置是否与es设定的安全值相匹配，如下：

• 如果匹配，则不会有warnning信息，此时es正常启动；
• 如果不匹配，则会有warnning信息，但因为是开发环境，es依然会正常启动；

1.2 生产环境

一旦用户配置了network.host来指定一个可用的非loopback地址，那么es就会认为用户此时是在生产环境下启动es，同样会进行检查，但一旦检查不通过，直接会将前面的warnning提升为error，所以此时es会启动失败。

2 开发环境启动时的bootstrap checks分析

不配置network.host时，直接启动es，会有下面的warnning：

[2018-12-07T04:15:44,735][INFO ][o.e.d.DiscoveryModule    ] [PQ85ukj] using discovery type [zen]
[2018-12-07T04:15:45,702][INFO ][o.e.n.Node               ] initialized
[2018-12-07T04:15:45,703][INFO ][o.e.n.Node               ] [PQ85ukj] starting ...
[2018-12-07T04:15:46,071][INFO ][o.e.t.TransportService   ] [PQ85ukj] publish_address {127.0.0.1:9300}, bound_addresses {[::1]:9300}, {127.0.0.1:9300}
[2018-12-07T04:15:46,090][WARN ][o.e.b.BootstrapChecks    ] [PQ85ukj] max file descriptors [4096] for elasticsearch process is too low, increase to at least [65536]
[2018-12-07T04:15:46,090][WARN ][o.e.b.BootstrapChecks    ] [PQ85ukj] max number of threads [1024] for user [hadoop] is too low, increase to at least [2048]
[2018-12-07T04:15:46,090][WARN ][o.e.b.BootstrapChecks    ] [PQ85ukj] max virtual memory areas vm.max_map_count [65530] is too low, increase to at least [262144]
[2018-12-07T04:15:46,090][WARN ][o.e.b.BootstrapChecks    ] [PQ85ukj] system call filters failed to install; check the logs and fix your configuration or disable system call filters at your own risk
[2018-12-07T04:15:49,269][INFO ][o.e.c.s.ClusterService   ] [PQ85ukj] new_master {PQ85ukj}{PQ85ukjdSoeVEpSpByAjMw}{Dbb3lzTWTN-eUEKXO8z-sw}{127.0.0.1}{127.0.0.1:9300}, reason: zen-disco-elected-as-master ([0] nodes joined)
[2018-12-07T04:15:49,313][INFO ][o.e.h.n.Netty4HttpServerTransport] [PQ85ukj] publish_address {127.0.0.1:9200}, bound_addresses {[::1]:9200}, {127.0.0.1:9200}
[2018-12-07T04:15:49,313][INFO ][o.e.n.Node               ] [PQ85ukj] started
[2018-12-07T04:15:49,553][INFO ][o.e.g.GatewayService     ] [PQ85ukj] recovered [0] indices into cluster_state

提取其waarnning信息，如下：

文件描述符：
max file descriptors [4096] for elasticsearch process is too low, increase to at least [65536]

线程数：
 max number of threads [1024] for user [hadoop] is too low, increase to at least [2048]

 虚拟内存：
 max virtual memory areas vm.max_map_count [65530] is too low, increase to at least [262144]

 system call filters:
 system call filters failed to install; check the logs and fix your configuration or disable system call filters at your own risk

可以看到有4个问题，分别为：文件描述符、线程数、虚拟内存与system call filters。

虽然有warnning，但因为es本身会认为是在开发环境下运行，基于开箱即用的特性，是仍然可以正常启动的。

3 生产环境启动时的bootstrap checks分析

绑定IP地址后再启动，发现有下面的报错信息：

ERROR: [4] bootstrap checks failed
[1]: max file descriptors [4096] for elasticsearch process is too low, increase to at least [65536]
[2]: max number of threads [1024] for user [hadoop] is too low, increase to at least [2048]
[3]: max virtual memory areas vm.max_map_count [65530] is too low, increase to at least [262144]
[4]: system call filters failed to install; check the logs and fix your configuration or disable system call filters at your own risk

直接error，所以会启动失败，除非进行上面的设置符合安全要求。

4 生产环境正常启动配置

解决上面出现的问题，需要进行如下的配置。

4.1 文件描述符

• 临时修改：

 ulimit -n 65536

但是重新登录后就会恢复成默认值了。

• 永久修改

修改/etc/security/limits.conf配置，如下：

hadoop          soft    nofile  65536   # soft表示为超过这个值就会有warnning
hadoop          hadr    nofile  100000  # hard则表示不能超过这个值

之后再重新登录，使用ulimit -n就可以进行验证。

4.2 线程数

修改/etc/security/limits.conf配置，如下：

hadoop          soft    nproc   2048
hadoop          hard    nproc   4096

实际上，该配置文件对于nproc的说明为进程数，而不是线程数：

#<domain>      <type>  <item>         <value>
                                                                                                   39,1          41%
# /etc/security/limits.conf
#
#Each line describes a limit for a user in the form:
#
#<domain>        <type>  <item>  <value>
#
#Where:
#<domain> can be:
#        - an user name
#        - a group name, with @group syntax
#        - the wildcard *, for default entry
#        - the wildcard %, can be also used with %group syntax,
#                 for maxlogin limit
#
#<type> can have the two values:
#        - "soft" for enforcing the soft limits
#        - "hard" for enforcing hard limits
#
#<item> can be one of the following:
#        - core - limits the core file size (KB)
#        - data - max data size (KB)
#        - fsize - maximum filesize (KB)
#        - memlock - max locked-in-memory address space (KB)
#        - nofile - max number of open files
#        - rss - max resident set size (KB)
#        - stack - max stack size (KB)
#        - cpu - max CPU time (MIN)
#        - nproc - max number of processes
#        - as - address space limit (KB)
#        - maxlogins - max number of logins for this user
#        - maxsyslogins - max number of logins on the system
#        - priority - the priority to run user process with
#        - locks - max number of file locks the user can hold
#        - sigpending - max number of pending signals
#        - msgqueue - max memory used by POSIX message queues (bytes)
#        - nice - max nice priority allowed to raise to values: [-20, 19]
#        - rtprio - max realtime priority
#
#<domain>      <type>  <item>         <value>
#

#*               soft    core            0
#*               hard    rss             10000

4.3 虚拟内存

• 查看当前值

sysctl vm.max_map_count

• 临时设置

sysctl -w vm.max_map_count=262144

但是重启系统后就会失效。

• 永久性设置

修改配置文件/etc/sysctl.conf，如下：

vm.max_map_count=262144

需要重启后才生效。

4.4 system call filters

• 原因
  这是在因为Centos6不支持SecComp，而ES5.4.0默认bootstrap.system_call_filter为true进行检测，所以导致检测失败，失败后直接导致ES不能启动。

• 解决
  在elasticsearch.yml中配置bootstrap.system_call_filter为false，注意要在Memory下面:
  bootstrap.memory_lock: false
  bootstrap.system_call_filter: false

参考：https://www.jianshu.com/p/89f8099a6d09

原文：http://blog.51cto.com/xpleaf/2327317