集群网卡物理地址

H3CVNFManager产品故障处理手册Copyright2016杭州华三通信技术有限公司版权所有,保留一切权利.
非经本公司书面许可,任何单位和个人不得擅自摘抄、复制本文档内容的部分或全部,并不得以任何形式传播.
本文档中的信息可能变动,恕不另行通知.
1-1目录1简介····························································································································1-31.
1故障处理注意事项·······································································································1-31.
2收集故障信息·············································································································1-41.
2.
1收集VNFManager诊断信息················································································1-41.
2.
2收集VNFManager日志信息················································································1-52VNFManager安装授权相关故障处理方法··········································································2-72.
1VNFManager安装时依赖检查故障处理方法·····································································2-72.
1.
1故障描述··········································································································2-72.
1.
2故障处理思路····································································································2-72.
1.
3故障处理流程····································································································2-82.
1.
4故障处理步骤····································································································2-82.
1.
5故障诊断命令··································································································2-102.
1.
6故障信息收集方法····························································································2-102.
2单台VNFManager或集群重启后无法获取远端License信息·············································2-102.
2.
1故障描述········································································································2-102.
2.
2故障处理思路··································································································2-102.
2.
3故障处理流程··································································································2-112.
2.
4故障处理步骤··································································································2-112.
2.
5故障诊断命令··································································································2-132.
2.
6故障信息收集方法····························································································2-133VNFManager集群故障处理方法·····················································································3-133.
1VNFManager创建集群或Region时失败报错·································································3-133.
1.
1故障描述········································································································3-133.
1.
2故障处理思路··································································································3-133.
1.
3故障处理流程··································································································3-143.
1.
4故障处理步骤··································································································3-143.
1.
5故障诊断命令··································································································3-153.
1.
6故障信息收集方法····························································································3-163.
2集群IP地址登录失败································································································3-163.
2.
1故障描述········································································································3-163.
2.
2故障处理思路··································································································3-163.
2.
3故障处理流程··································································································3-161-23.
2.
4故障处理步骤··································································································3-173.
2.
5故障诊断命令··································································································3-183.
2.
6故障信息收集方法····························································································3-183.
3无法清除MemberVNFManager上的集群配置·······························································3-183.
3.
1故障描述········································································································3-183.
3.
2故障处理思路··································································································3-183.
3.
3故障处理流程··································································································3-193.
3.
4故障处理步骤··································································································3-193.
3.
5故障诊断命令··································································································3-203.
3.
6故障信息收集方法····························································································3-203.
4VNFManager无法成功加入集群·················································································3-203.
4.
1故障描述········································································································3-203.
4.
2故障处理思路··································································································3-203.
4.
3故障处理流程··································································································3-213.
4.
4故障处理步骤··································································································3-213.
4.
5故障诊断命令··································································································3-223.
4.
6故障信息收集方法····························································································3-223.
5成员VNFManager显示状态为down············································································3-233.
5.
1故障描述········································································································3-233.
5.
2故障处理思路··································································································3-233.
5.
3故障处理流程··································································································3-233.
5.
4故障处理步骤··································································································3-233.
5.
5故障诊断命令··································································································3-243.
5.
6故障信息收集方法····························································································3-243.
6集群主备成员数据不同步····························································································3-243.
6.
1故障描述········································································································3-243.
6.
2故障处理思路··································································································3-253.
6.
3故障处理流程··································································································3-253.
6.
4故障处理步骤··································································································3-263.
6.
5故障诊断命令··································································································3-263.
6.
6故障信息收集方法····························································································3-274诊断信息及NETCONF故障处理方法················································································4-274.
1导出集群诊断信息时,部分VNFManager导出异常·························································4-274.
1.
1故障描述········································································································4-274.
1.
2故障处理思路··································································································4-274.
1.
3故障处理流程··································································································4-271-34.
1.
4故障处理步骤··································································································4-284.
1.
5故障诊断命令··································································································4-284.
1.
6故障信息收集方法····························································································4-284.
2虚机的运行数据无法显示····························································································4-284.
2.
1故障描述········································································································4-284.
2.
2故障处理思路··································································································4-284.
2.
3故障处理流程··································································································4-294.
2.
4故障处理步骤··································································································4-294.
2.
5故障诊断命令··································································································4-294.
2.
6故障信息收集方法····························································································4-305VNFManager功能故障处理方法·····················································································5-305.
1VNFManager无法成功创建VNF或vBRAS状态异常······················································5-305.
1.
1故障描述········································································································5-305.
1.
2故障处理思路··································································································5-305.
1.
3故障处理流程··································································································5-315.
1.
4故障处理步骤··································································································5-315.
1.
5故障诊断命令··································································································5-395.
1.
6故障信息收集方法····························································································5-391简介本文档介绍H3CVNFManager(简称VNFManager)常见故障的诊断及处理措施.
1.
1故障处理注意事项当出现故障时,请尽可能全面、详细地记录现场信息(包括但不限于以下内容),收集信息越全面、越详细,越有利于故障的快速定位.
记录您所使用的H3CVNFManager版本、Linux操作系统版本.
记录具体的故障现象、故障时间、配置信息.
记录完整的网络拓扑,包括组网图、端口连接关系、故障位置.
收集日志信息和诊断信息(收集方法见1.
2收集故障信息).
记录现场采取的故障处理措施及实施后的现象效果.
1-41.
2收集故障信息1.
2.
1收集VNFManager诊断信息您可以通过如下步骤,查看VNFManager的诊断信息.
(1)在浏览器(如Chrome)中输入VNFManager的GUI的登录地址(格式为:http://http://192.
168.
152.
206/sdn/ui/app/index),回车后打开VNFManager的GUI登录界面.
输入用户名和密码后,单击按钮进入VNFManager的GUI首页.
(2)在GUI界面中,单击[系统/系统管理/诊断信息]菜单项,进入诊断信息页面,如图1-2-1所示.
然后单击右上角的按钮,弹出"诊断信息导出"对话框,如图1-2-2所示.
在弹出的对话框中选择需要导出诊断信息的VNFManager,单击按钮,将导出的诊断日志信息保存到本地.
集群模式下,当前LeaderVNFManager有权限导出集群内所有处于激活状态的设备诊断信息,而MemberVNFManager仅能导出本设备的诊断信息.
在已弹出的"诊断信息导出"对话框中,每台VNFManager的诊断信息仅能成功导出一次,如需对同一台VNFManager的诊断信息再次导出,请关闭对话框后重新打开.
1-5图1-2-1诊断信息页面图1-2-2"诊断信息导出"对话框1.
2.
2收集VNFManager日志信息(1)收集VNFManager操作日志:在GUI界面中,单击[运维监控/操作日志]菜单项,进入操作日志页面,如图1-2-3所示.
然后单击右上角的按钮,将导出的操作日志信息保存到本地.
1-6图1-2-3"操作日志导出"对话框(2)收集VNFManager系统日志:在GUI界面中,单击[运维监控/系统日志]菜单项,进入操作日志页面,如图1-2-4所示.
然后单击右上角的按钮,将导出的系统日志信息保存到本地.
图1-2-4"系统日志导出"对话框(3)收集VNFManager后台日志:登录VNFManager所在操作系统命令行界面,导出后台日志,路径:/opt/sdn/virgo/serviceability/logs/(4)收集Openstack后台日志:收集Openstack日志可使用RemoteExportOpsLogs.
jar小工具获取.
如下图所示,在小工具UI节目输入对应节点的IP地址、用户名、密码,设置好本地下载路径,点击"ExportLogs"按钮,即可将不正常节点的Openstack日志下载到本地对应路径.
本例中不正常节点地址为"99.
1.
1.
131".
2-7图1-2-52VNFManager安装授权相关故障处理方法2.
1VNFManager安装时依赖检查故障处理方法2.
1.
1故障描述安装VNFManager时,提示"PleaseinstallJRE-headlessjava-1.
8.
0-openjdk-amd64orjava-1.
7.
0-openjdk-amd64beforeyouinstallthecontroller.
"报错信息,如下图所示.
图2-1-1安装VNFManager时的报错信息2.
1.
2故障处理思路造成故障的原因可能为未正确安装VNFManager依赖的OpenJDK软件包.
2-82.
1.
3故障处理流程2.
1.
4故障处理步骤1.
检查离线依赖包安装版本使用是否正确使用命令cat/etc/redhat-release查看使用的操作系统与安装的JDK升级包版本是否一致.
2.
检查依赖包安装是否正确(1)安装依赖包的过程中会自动执行关闭防火墙的命令.
通过systemctlstatusfirewalld.
service命令检查防火墙是否关闭.
2-9[root@localhost~]#systemctlstatusfirewalld.
servicefirewalld.
service-firewalld-dynamicfirewalldaemonLoaded:loaded(/usr/lib/systemd/system/firewalld.
service;disabled)Active:inactive(dead)(2)通过virsh–version命令验证Libvirt软件包是否安装成功.
[root@localhost~]#virsh-version1.
1.
1(3)通过systemctlstatuspostgresql命令验证PostgreSQL软件是否安装和启动成功.
[root@localhost~]#systemctlstatuspostgresqlpostgresql.
service-PostgreSQLdatabaseServerLoaded:loaded(/usr/lib/systemd/system/postgresql.
service;enabled)Active:active(running)sinceFri2016-03-2515:00:31CST;31minago…略…(4)通过/usr/lib/jvm/java-1.
8.
0/bin/java-version命令验证OpenJDK1.
8.
0软件包是否安装成功.
[root@localhost~]#/usr/lib/jvm/java-1.
8.
0/bin/java-versionopenjdkversion"1.
8.
0_65"OpenJDKRuntimeEnvironment(build1.
8.
0_65-b17)OpenJDK64-BitServerVM(build25.
65-b01,mixedmode)(5)通过lsb_release--help命令验证Redhat-lsb-core软件包是否安装成功.
[root@localhostjvm]#lsb_release--helpFSGlsb_releasev2.
0printscertainLSB(LinuxStandardBase)andDistributioninformation.
…略…(6)通过unzip-v命令验证Unzip软件包是否安装成功.
[root@localhostjvm]#unzip-vUnZip6.
00of20April2009,byInfo-ZIP.
MaintainedbyC.
Spieler.
…略…3.
安装正确的依赖包软件建议通过安装H3C提供的在线安装依赖软件包的方式安装VNFManager依赖的所有软件包.
在线安装VNFManager依赖软件包的方法,请参见《H3CVNFManager安装指导-CentOS》.
如果需要通过离线依赖包的形式进行安装,请参加《H3CVNFManager离线依赖包安装指导-CentOS》.
离线依赖包安装完成后,可以通过rpm-qa|grepjdk查看jdk离线依赖包是否升级安装完成:[root@localhostlogs]#rpm-qa|grepjdkjava-1.
8.
0-openjdk-devel-1.
8.
0.
65-3.
b17.
el7.
x86_64java-1.
7.
0-openjdk-headless-1.
7.
0.
75-2.
5.
4.
2.
el7_0.
x86_64java-1.
8.
0-openjdk-headless-1.
8.
0.
65-3.
b17.
el7.
x86_64java-1.
7.
0-openjdk-javadoc-1.
7.
0.
75-2.
5.
4.
2.
el7_0.
noarchjava-1.
8.
0-openjdk-javadoc-1.
8.
0.
65-3.
b17.
el7.
noarchjava-1.
7.
0-openjdk-1.
7.
0.
75-2.
5.
4.
2.
el7_0.
x86_64java-1.
8.
0-openjdk-1.
8.
0.
65-3.
b17.
el7.
x86_64java-1.
7.
0-openjdk-devel-1.
7.
0.
75-2.
5.
4.
2.
el7_0.
x86_642-104.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
2.
1.
5故障诊断命令命令说明cat/etc/redhat-release显示Linux版本rpm-qa|grepjdk查看JDK依赖包安装情况systemctlstatusfirewalld.
service查看防火墙服务运行情况systemctlstatuspostgresql验证PostgreSQL软件是否安装和启动成功unzip-v验证Unzip软件包是否安装成功2.
1.
6故障信息收集方法参考章节1.
2《收集故障信息》2.
2单台VNFManager或集群重启后无法获取远端License信息2.
2.
1故障描述单VNFManager或集群中所有VNFManager全部重启后,与LicenseServer成功建立连接,并且在LicenseServer上显示License已安装,但VNFManager或集群获取不到远端License的授权信息.
2.
2.
2故障处理思路造成故障的原因可能为VNFManager或集群重启后,与LicenseServer的连接中断,而LicenseServer上会等待老化时间结束后再回收授权数据.
在老化时间结束前,即使与LicenseServer成功建立连接也不会获取到授权信息.
2-112.
2.
3故障处理流程2.
2.
4故障处理步骤1.
检查VNFManager是否重启过登录VNFManager后台SHELL界面,查看当前系统进程运行时间是否为近期.
使用命令servicesdnastatus查看当前sdna进程运行时间:[root@localhost~]#servicesdnastatussdna.
service-sdnasystemdconfLoaded:loaded(/etc/systemd/system/sdna.
service;enabled)Active:active(running)sinceFri2017-05-0522:05:55EDT;2h14minago2-12Process:12248ExecStart=/etc/systemd/system/startSdna(code=exited,status=0/SUCCESS)MainPID:12249(java)使用命令servicesdncstatus查看当前sdnc进程运行时间:sdnc.
service-sdncsystemdconfLoaded:loaded(/etc/systemd/system/sdnc.
service;enabled)Active:active(running)sinceFri2017-05-0522:35:30EDT;1h46minagoProcess:15267ExecStopPost=/usr/bin/sleep2(code=exited,status=0/SUCCESS)Process:15270ExecStart=/etc/systemd/system/startSdnc(code=exited,status=0/SUCCESS)MainPID:15292(java)2.
LicenseServer上强制下线客户端在浏览器中输入LicenseServer的GUI登录地址(格式为:http://lics_ip_address:8090/licsmanager/,如http://172.
16.
0.
227:8090/licsmanager/),回车后会登录到LicenseServer的管理界面.
在客户端管理的连接管理页面中将对应的客户端强制下线.
图2-2-1LicenseServer界面3.
VNFManager上重连LicenseServer登录到VNFManager,单击[系统/系统License]菜单项,在远端License页面断开与LicenseServer的连接,并重新建立连接.
图2-2-2VNFManager界面4.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3-132.
2.
5故障诊断命令命令说明servicesdnastatus查看sdna进程状态servicesdncstatus查看sdnc进程状态2.
2.
6故障信息收集方法参考章节1.
2《收集故障信息》3VNFManager集群故障处理方法3.
1VNFManager创建集群或Region时失败报错3.
1.
1故障描述创建集群时,失败报错.
提示"网卡配置不正确,集群创建失败"或"集群创建失败,集群成员集群Token不一致"或"不能在当前角色为Member的集群成员VNFManager上创建集群".
创建Region时,提示某台VNFManager配置失败.
3.
1.
2故障处理思路创建集群或Region失败的时候,对于不同的报错有不同的处理方法,要善于利用系统报错信息进行排障.
3-143.
1.
3故障处理流程3.
1.
4故障处理步骤1.
查看系统具体的报错信息创建集群或Region时,查看具体是哪一种报错报错:"网卡配置不正确,集群创建失败"或"集群创建失败,集群成员集群Token不一致"或"不能在当前角色为Member的集群成员VNFManager上创建集群".
2.
提示"网卡配置不正确,集群创建失败处理步骤"(1)检查VNFManager所在服务器或者虚拟机的网卡是否正常,如果网卡已禁用,请通过命令行启动网卡;如果是硬件故障,请更换网卡,然后重新创建集群.
(2)如果是网卡输入错误,请从"增加VNFManager"页的VNFManager列表中删除网卡错误的VNFManager,重新添加VNFManager并通过下拉框方式选择正确网卡,然后重新创建集群.
(3)如果通过下拉框方式无法显示远端LeaderVNFManager的网卡列表,请检查网络是否连通,如果网络连通正常,则请进一步检查各LeaderVNFManager的集群Token是否一致,查看集群Token的方法为:a.
登录VNFManager的GUI界面;b.
单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面;3-15c.
在单机配置页面可以查看当前VNFManager的集群Token信息.
(4)如果各LeaderVNFManager的集群Token不一致,需要通过如下方法修改集群Token后再重新创建集群.
方法一:登录VNFManager的GUI界面后,单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面,如果当前集群Token未创建,则在集群Token输入框中输入集群Token后单击按钮;如果当前集群Token已创建,则单击按钮删除当前集群Token,然后在输入框中输入新的集群Token,再单击按钮.
方法二:将需要修改集群Token的VNFManager卸载后重装,在安装过程提示输入集群Token时,输入与其它LeaderVNFManager相同的集群Token.
卸载和重装VNF的方法请参见《H3CVNFManager安装指导》.
3.
提示"集群创建失败,集群成员集群Token不一致"该问题是由于创建集群的各LeaderVNFManager的集群Token不一致造成的.
解决方法如下:(1)查看集群中各LeaderVNFManager的集群Token,查看方法为:a.
登录VNFManager的GUI界面;b.
单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面;c.
在单机配置页面中可以查看当前VNFManager的集群Token信息.
(2)修改LeaderVNFManager的集群Token,使得各LeaderVNFManager的集群Token一致.
修改方法为:方法一:登录VNFManager的GUI界面后,单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面,如果当前集群Token未创建,则在集群Token输入框中输入集群Token后单击按钮;如果当前集群Token已创建,则单击集群Token输入框后的按钮删除当前集群Token,然后在集群Token输入框中输入新的集群Token,再单击按钮.
方法二:将需要修改集群Token的VNFManager卸载后重装,在安装过程提示输入集群Token时,输入与其它LeaderVNFManager相同的集群Token.
卸载和重装VNFManager的方法请参见《H3CVNFManager安装指导-CentOS》.
(3)待集群Token修改完毕后,重新创建集群.
4.
提示"不能在当前角色为Member的集群成员VNFManager上创建集群"请将当前VNFManager设置为Leader角色,或者登录到角色为Leader的VNFManager上创建集群.
5.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
1.
5故障诊断命令命令说明不涉及不涉及3-163.
1.
6故障信息收集方法参考章节1.
2《收集故障信息》3.
2集群IP地址登录失败3.
2.
1故障描述无法通过集群IP地址登录集群,或ping不通集群IP地址.
3.
2.
2故障处理思路造成故障的原因可能为PC无法连接到VNFManager集群,或者集群IP相关配置错误,或者PC端ARP信息错误问题.
3.
2.
3故障处理流程3-173.
2.
4故障处理步骤1.
检查VNFManager集群Leader和PC之间连通性在PC上ping当前Leader的IP地址,如果无法ping通,请排查PC和当前Leader所属网段之间的网络连通性.
2.
检查当前LeaderVNFManager上是否存在集群IP地址检查当前LeaderVNFManager上是否存在集群IP地址,可以通过单击[系统/系统信息]或[运维监控/系统信息]菜单项查看.
如果不存在集群IP地址,请单击[系统/系统配置]菜单项,然后单击"修改集群"链接,进入修改集群页面.
再单击按钮,在弹出的对话框中填写集群IP地址和子网掩码,单击按钮来创建集群IP地址.
图3-2-1修改集群界面3.
VNFManager上重连LicenseServer登录当前LeaderVNFManager所在的服务器或者虚拟机,使用ifconfig命令查看设置的网卡上是否存在集群IP地址.
如果无集群IP地址,请检查网卡是否禁用,如果已禁用,请使用ifconfig命令启动网卡.
如下所示,物理网卡地址为99.
1.
1.
132,集群IP为99.
1.
1.
134.
[root@localhostlogs]#ifconfigeno16777728:flags=4163mtu1500inet99.
1.
1.
132netmask255.
255.
255.
0broadcast99.
1.
1.
255inet6fe80::250:56ff:fe97:2ea6prefixlen64scopeid0x20ether00:50:56:97:2e:a6txqueuelen1000(Ethernet)RXpackets198236172bytes26608538143(24.
7GiB)RXerrors0dropped88732overruns0frame0TXpackets184176955bytes20488341928(19.
0GiB)TXerrors0dropped0overruns0carrier0collisions03-18eno16777728:0:flags=4163mtu1500inet99.
1.
1.
134netmask255.
255.
255.
255broadcast0.
0.
0.
0ether00:50:56:97:2e:a6txqueuelen1000(Ethernet)4.
检查网络中是否存在集群IP地址冲突检查网络中是否存在集群IP地址冲突,如果存在,请登录当前LeaderVNFManager,把集群IP地址修改为当前网络中唯一的IP地址.
5.
PC上查看ARP表项在浏览器所在主机,也就是登录VNFManager的主机上查看ARP表项,确认集群IP地址所对应的MAC地址是否为当前LeaderVNFManager上集群IP地址对应网卡的MAC地址.
如果不是,则表明ARP表项不正确,请删除ARP表项,然后重新ping集群IP地址.
6.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
2.
5故障诊断命令命令说明ifconfig查看Linux网卡IP配置3.
2.
6故障信息收集方法参考章节1.
2《收集故障信息》3.
3无法清除MemberVNFManager上的集群配置3.
3.
1故障描述无法清除MemberVNFManager上的集群配置.
3.
3.
2故障处理思路由于网络不可达,导致成员VNFManager无法和LeaderVNFManager交互信息,或者在删除集群时,由于成员VNFManager下线等原因导致未同步删除集群配置信息,都有可能导致使得成员VNFManager无法清除集群配置信息.
3-193.
3.
3故障处理流程3.
3.
4故障处理步骤只有运行角色为Leader的VNFManager才有删除集群的权限.
当网络中不存在具备删除权限的VNFManager用于清理集群配置时,就会导致MemberVNFManager残留集群配置.
1.
检查VNFManager集群是否已经删除如果集群未删除,请确保MemberVNFManager与LeaderVNFManager之间网络可达后再尝试删除集群.
2.
通过MemberVNFManager逐一退出集群的方式清除集群配置如果集群已删除,请通过MemberVNFManager逐一退出集群的方式清除集群配置.
操作方法为:(1)登录MemberVNFManager的GUI界面.
单击[系统/系统配置]菜单项,进入系统配置页面.
(2)单击"修改集群"链接,进入修改集群页面,在VNFManager列表中,单击所登录VNFManager对应"操作"列中的按钮.
(3)在弹出的操作确认对话框中单击按钮,完成所登录的VNFManager退出集群的操作.
(4)重复上述a~c步骤,使所有残留集群配置的MemberVNFManager逐一退出集群.
3-20图3-3-1修改集群界面3.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
3.
5故障诊断命令命令说明不涉及不涉及3.
3.
6故障信息收集方法参考章节1.
2《收集故障信息》3.
4VNFManager无法成功加入集群3.
4.
1故障描述向集群中动态添加集群成员时,成员VNFManager无法成功加入.
3.
4.
2故障处理思路由于网络不可达,导致成员VNFManager无法和LeaderVNFManager交互信息,或者集群成员数量达到上限,或者集群Token不一致,都有可能导致VNFManager集群成员无法加入.
3-213.
4.
3故障处理流程3.
4.
4故障处理步骤1.
检查成员VNFManager和当前LeaderVNFManager之间HTTPS连接是否正常检查成员VNFManager和当前LeaderVNFManager之间HTTPS连接是否正常,可以在成员VNFManager上尝试ping集群IP地址,如果无法ping通,建议排查底层连通性.
2.
检查成员VNFManager配置是否存在错误检查成员VNFManager配置是否存在错误,例如IP地址或名称与集群内已有成员是否重复.
如果是成员VNFManager的IP地址、名称重复等错误,请根据提示修改相应配置信息.
3-22图3-4-1修改集群界面3.
查看集群成员数量查看集群内成员数目是否已经达到规格上限,如果已经满规格,则无法加入.
请重新规划集群成员,只有将原有集群成员退出后才能重新加入新成员.
4.
检查待加入集群的VNFManager和集群当前LeaderVNFManager的集群Token是否一致检查待加入集群的VNFManager和集群当前LeaderVNFManager的集群Token是否一致,如果不一致,请修改待加入集群的VNFManager的集群Token,具体方法如下:方法一:登录VNFManager的GUI界面后,单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面,如果当前集群Token未创建,则在集群Token输入框中输入集群Token后单击按钮;如果当前集群Token已创建,则单击按钮删除当前集群Token,然后在输入框中输入新的集群Token,再单击按钮.
方法二:将需要修改集群Token的VNFManager卸载后重装,在安装过程提示输入集群Token时,输入新的集群Token.
卸载和重装的方法请参见《H3CVNFManager安装指导》.
5.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
4.
5故障诊断命令命令说明不涉及不涉及3.
4.
6故障信息收集方法参考章节1.
2《收集故障信息》3-233.
5成员VNFManager显示状态为down3.
5.
1故障描述查看成员VNFManager的状态信息,显示为down.
3.
5.
2故障处理思路由于网络中断或VNFManager宕机,或者VNFManager同时属于两个集群,或者集群Token配置错误,都有可能导致VNFManager状态信息为down.
3.
5.
3故障处理流程3.
5.
4故障处理步骤1.
检查成员VNFManager是否宕机或网络是否中断在VNFManager信息页面中,查看VNFManager信息列表备注栏的提示,如果显示"VNFManagerIP不可达"或"连接超时",请检查VNFManager是否宕机或网络是否中断.
如果VNF3-24Manager宕机请重新上电或者重新安装VNFManager;如果网络故障,请排查故障线路,确保网络可达.
2.
检查成员VNFManager是否同时存在于两个集群在VNFManager信息页面中,查看VNFManager信息列表备注栏的提示,如果显示"该VNFManager同时存在于两个集群中",请根据需求将VNFManager退出之前加入或新加入的集群.
3.
查看成员VNFManager集群Token配置是否正确在VNFManager信息页面中,查看VNFManager信息列表备注栏的提示,如果显示"集群Token认证失败",请通过如下方法解决:查看集群中各LeaderVNFManager的集群Token,查看方法为:登录VNFManager的GUI界面后,单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面,在单机配置页面中可以查看当前VNFManager的集群Token信息.
修改VNFManager的集群Token,使得各VNFManager的集群Token一致.
修改方法为:方法一:登录VNFManager的GUI界面后,单击[系统/系统配置]菜单项,然后单击"单机配置"链接进入单机配置页面,如果当前集群Token未创建,则在输入框中输入集群Token后单击按钮;如果当前集群Token已创建,则单击输入框后的按钮删除当前集群Token,然后再输入新的集群Token,再单击按钮.
方法二:将需要修改集群Token的VNFManager卸载后重装,在安装过程提示输入集群Token时,输入与其它LeaderVNFManager相同的集群Token.
卸载和重装VNFManager的方法请参见《H3CVNFManager安装指导》.
修改集群Token后,重新创建集群或者将VNFManager加入集群.
4.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
5.
5故障诊断命令命令说明不涉及不涉及3.
5.
6故障信息收集方法参考章节1.
2《收集故障信息》3.
6集群主备成员数据不同步3.
6.
1故障描述查看登录集群主Leader成员和其他全部成员查看VNF运行相关信息,显示不一致.
如图3-6-1所示.
3-25图3-6-1主备成员数据不同步3.
6.
2故障处理思路集群主备成员数据不同步,可以先查看成员状态是否正常,如果不正常,根据VNFManager备注提示信息进行修复.
如果修复后还无法同步数据,可重启VNFManager应用强制数据同步.
3.
6.
3故障处理流程3-263.
6.
4故障处理步骤1.
检查成员VNFManager状态是否正常登陆LeaderVNFManager,进入"系统/系统信息"界面,查看集群状态信息,如下图,红色圈出的为备注信息.
图3-6-2检查成员状态2.
解决成员VNFManager备注信息异常问题备注信息处理方法如下:连接超时:Down状态的VNFManager发生了宕机或者网络中断,请查看对应成员所在的服务器状态及排除网络故障.
VNFManagerIP不可达:Down状态的VNFManager可能发生宕机、网络中断、未安装VNFManager软件或安装完VNFManager软件后未完成启动.
网络中断需要检查成员VNFManager所在服务器或虚拟机是否存在VNFManager对应的IP地址,如果不存在或者被禁用,则正确配置VNFManagerIP地址后重启VNFManager.
集群token认证失败:重新升级VNFManager,且更新Token配置.
若数据仍不同步,重启LeaderVNFManagerH3CVNFManager应用.
3.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
3.
6.
5故障诊断命令命令说明不涉及不涉及4-273.
6.
6故障信息收集方法首先参考章节1.
2《收集故障信息》收集主/备VNFManager诊断、日志、后台日志等信息.
然后收集主/备VNFManager:Orchestrator数据库备份文件,请在VNFManager后台执行pg_dumpsdndb>sdndb.
sql命令,在root目录生成sdndb.
sql文件,后台导出.
4诊断信息及NETCONF故障处理方法4.
1导出集群诊断信息时,部分VNFManager导出异常4.
1.
1故障描述导出集群内VNFManager诊断信息时,VNFManager诊断信息未成功导出或者导出的压缩包不能正常进行解压.
4.
1.
2故障处理思路造成故障的原因可能为:受CPU过于繁忙、网络不稳定、日志过大等因素的影响,在导出集群诊断信息的时候与某台VNFManager连接超时,导致无法正常导出该VNFManager的诊断信息.
4.
1.
3故障处理流程4-284.
1.
4故障处理步骤1.
查看系统具体的报错信息创建集群或Region时,查看具体是哪一种报错报错:"网卡配置不正确,集群创建失败"或"集群创建失败,集群成员集群Token不一致"或"不能在当前角色为Member的集群成员VNFManager上创建集群".
2.
尝试重新导出异常VNFManager诊断登录出现异常的VNFManager的GUI界面,在该VNFManager上再次导出诊断信息.
3.
在Leader上导出异常VNFManager诊断在当前LeaderVNFManager的诊断信息页面上再次单击按钮,在弹出的"诊断信息导出"对话框中选择出现异常的VNFManager,导出诊断信息.
4.
后台登录异常VNFManager后台日志通过SSH登录到出现异常的VNFManager所在的服务器或虚拟机,进入/opt/sdn/virgo/serviceability/logs目录下,使用FTP等文件传输工具获取需要的后台日志信息.
5.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
4.
1.
5故障诊断命令命令说明不涉及不涉及4.
1.
6故障信息收集方法参考章节1.
2《收集故障信息》4.
2虚机的运行数据无法显示4.
2.
1故障描述在虚机摘要信息页面,虚机设备的部分信息显示为"---".
4.
2.
2故障处理思路造成故障的原因可能为:虚机状态异常、虚机NETCONF配置错误、虚机NETCONF规格达到上限.
4-294.
2.
3故障处理流程4.
2.
4故障处理步骤1.
确认虚机与VNFManager通信是否正常确认虚机设备与VNFManager通信是否正常.
分别登录VNFManager和虚机设备,通过ping命令查看是否可以正常通信.
如果不正常,排查底层连通性设置2.
确认虚机与VNFManagerNETCONF连接配置是否一致确认虚机设备和VNFManager的NETCONF相关配置是否一致.
如虚机设备是否开启了基于HTTPS的NETCONFoverSOAP功能;VNFManager使用的NETCONF用户名和密码与虚机设备上的NETCONF用户名和密码是否一致.
若不一致,请修改虚机设备或VNFManager的NETCONF配置.
3.
确认VNF设备和其它设备建立的NETCONF会话数是否已达上限确认VNF设备和其它设备建立的NETCONF会话数是否已达上限.
因为设备能够建立的NETCONF会话数有限,如果虚机设备和其它设备建立的NETCONF会话已达上限,就无法再与VNFManager建立会话,此时需要删除虚机设备和其它设备建立的NETCONF会话或调整网络设备的NETCONF会话上限.
4.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,收集信息拨打H3C热线400-810-0504寻求帮助.
4.
2.
5故障诊断命令命令说明不涉及不涉及5-304.
2.
6故障信息收集方法参考章节1.
2《收集故障信息》5VNFManager功能故障处理方法5.
1VNFManager无法成功创建VNF或vBRAS状态异常5.
1.
1故障描述VNFManager创建VNF后,点击"完成"按钮后仍然无法正常创建出VNF:即在VNF管理/VNF运行信息一栏看不到自己创建的VNF或者VNF状态不为"Active".
5.
1.
2故障处理思路处理此类故障时,首先要排查Openstack节点安装是否正确、License是否激活,是否设置有集群Token,是否能正常添加主机且主机状态为Normal,保证可以成功开始创建VNF.
如果开始创建VNF之后报错或创建VNF状态ERROR,需要保证主机节点的时间同步性、资源的充足性.
5-315.
1.
3故障处理流程5.
1.
4故障处理步骤1.
主机状态为Abnormal问题现象如图5-1-1所示.
Host主机列表,主机状态显示Abnormal.
5-32图5-1-2可能原因为未正常安装Openstack控制或计算节点,建议参考《VNFManager软件产品开局指导》检查安装情况.
2.
查看控制节点与计算节点系统时间是否一致登录到OpenStack控制节点和计算节点主机的控制台Linux命令行界面,用date命令查看系统时间,如果系统时间不一致可能导致VNF创建失败,需要手动修改控制节点和计算节点主机系统时间一致或配置NTP服务器.
(1)查看OpenStack控制节点和计算节点主机的系统时间是否一致登录到OpenStack控制节点和计算节点主机的控制台Linux命令行界面,登录之后使用"data"命令,分别查看各个节点的系统时间,即本例中标红加粗部分,保证控制节点与计算节点互相之间系统时间的差值在一分钟以内:(2)调整Openstack控制节点和计算节点时间调整Openstack控制节点和计算节点主机系统时间,可采用配置NTP时间服务器或手动调整的方式.
本文就手动调整方式进行介绍.
首先,通过步骤(1)确定控制节点时间,本例中控制节点时间为"TueJan3117:21:34CST2017".
完成确认后,登录计算节点控制台命令行界面,通过"date-s"命令配置计算节点系统时间和控制节点系统时间一致.
然后,通过"clock-w"命令将计算节点系统时间同步到计算节点主机的硬件时间,保证计算节点主机重启后系统时间不会丢失.
通过"clock"命令可查看当前的硬件时间是否已经和系统时间保持一致,本例中标红加粗部分即为主机硬件时间.
最后通过"reboot"命令重启计算节点主机Linux系统.
[root@ctrl1~]#dateTueJan3117:21:34CST2017[root@compute1~]#date-s"TueJan3117:21:34CST2017"TueJan3117:21:34CST20175-333.
查看控制节点资源配额余量是否满足VNFFlavor查看Openstack控制节点租户配额的CPU内核数、内存、Servergroups剩余的数量是否满足VNF创建使用的flavor大小,如果不满足,需要考虑扩展租户的资源配额.
(1)查看Openstack控制节点的租户名称登录到OpenStack控制节点的控制台命令行界面,登录之后使用"keystonetenant-list"命令,查看控制节点上存在哪些租户,即本例中标红加粗部分,即"admin"(租户ID为26975ace82564da3b24f873754a9cb94)与"services"(租户ID为cbc9a1caf15a42b98807c60a0716b384).
在使用VNFManager配套的kilo.
iso镜像部署完控制节点后,默认会生成以上两个租户,通常默认使用admin租户创建资源.
(2)查询控制节点租户的资源配额登录到OpenStack控制节点的控制台命令行界面,登录之后使用"novaquota-show"命令,查看控制节点上租户的资源配额.
主要查看CPU内核数、内存、Servergroups的配额值,即本例中标红加粗部分:可以看到租户最多可以创建25个VNF实例(instance),VNF的CPU内核数(cores)总和不能超过250个,内存总使用量(ram)上限为204800M,可以创建的VNF堆叠组(server_groups)数量不能超过20个.
[root@compute1~]#clock–w[root@compute1~]#clockTue31Jan201705:21:34PMCST-0.
641119seconds[root@compute1~]#reboot[root@ctrl1~]#keystonetenant-list/usr/lib/python2.
7/site-packages/keystoneclient/shell.
py:65:DeprecationWarning:ThekeystoneCLIisdeprecatedinfavorofpython-openstackclient.
ForaPythonlibrary,continueusingpython-keystoneclient.
'python-keystoneclient.
',DeprecationWarning)|id|name|enabled||26975ace82564da3b24f873754a9cb94|admin|True||cbc9a1caf15a42b98807c60a0716b384|services|True|5-34(3)查询控制节点特定租户的资源使用情况登录到OpenStack控制节点的控制台命令行界面,登录之后使用"mysql"命令进入安装部署kilo.
iso时自动部署的数据库,默认无密码.
使用"usenova"命令进入nova数据库,看到标红加粗字体Databasechanged时说明成功进入数据库.
进入nova数据库后,使用"select*fromquota_usages"命令查看目前各租户的资源使用情况.
如下图所示,通过标红加粗字体部分可以看到租户"26975ace82564da3b24f873754a9cb94"即租户"admin"目前已使用的实例数量为13个,使用的内存为16384M,使用的vCPU核数位22个.
[root@ctrl1~]#novaquota-show|Quota|Limit||instances|25||cores|250||ram|204800||floating_ips|50||fixed_ips|-1||metadata_items|256||injected_files|2||injected_file_content_bytes|10240||injected_file_path_bytes|255||key_pairs|10||security_groups|15||security_group_rules|20||server_groups|20||server_group_members|10|[root@ctrl1~]#mysqlWelcometotheMariaDBmonitor.
Commandsendwith;or\g.
YourMariaDBconnectionidis3939Serverversion:5.
5.
40-MariaDB-wsrepMariaDBServer,wsrep_25.
11.
r4026Copyright(c)2000,2014,Oracle,MariaDBCorporationAbandothers.
Type'help;'or'\h'forhelp.
Type'\c'toclearthecurrentinputstatement.
8rowsinset(0.
01sec)MariaDB[(none)]>usenovaReadingtableinformationforcompletionoftableandcolumnnamesYoucanturnoffthisfeaturetogetaquickerstartupwith–ADatabasechanged5-35结合之前查看到的租户"admin"的配额值,我们可以判断目前租户"admin"使用的资源还未超出他的配额,可以继续使用该租户创建VNF.
(4)调整租户的资源配额如果租户使用资源超出配额,可手动调整特定租户的配额.
登录到OpenStack控制节点的控制台命令行界面,登录之后使用"novaquota-update"命令即可动态调整特定租户ID的资源配额.
如下所示,本例中将租户"26975ace82564da3b24f873754a9cb94"即租户"admin"可创建的实例数调整为30,然后通过之后使用"novaquota-show"命令,查看控制节点上租户的资源配额,可看到可创建VNF数"instances"已调整为"30".
5-364.
查看Openstack控制节点和计算节点的网络映射关系配置是否一致查看Openstack控制节点和计算节点的网络映射关系配置是否一致,保证控制节点的物理网络在计算节点上有配置对应的物理网络和网卡的映射关系,如果不一致,手动修改计算节点上物理网络和网卡的映射关系.
(1)查看Openstack控制节点创建VNF使用的逻辑网络与物理网络的映射关系登录到OpenStack控制节点的控制台命令行界面,登录之后执行neutronnet-show{name|id}命令查看指定逻辑网络与物理网络的映射关系,其中,name字段是逻辑网络名称,本例中为标红加粗字段"network1","provider:physical_network"后接字段是物理网络名称,可以看到逻辑网络"network1"对应的物理网络为"provider1".
这里需要注意:如果控制节点上的逻辑网络最终映射到计算节点的SR-IOV类型网卡,逻辑网络名称即name字段必须以SRIOV作为前缀,例如SRIOV_ens1f0、SRIOV_ens1f1、SRIOV_0、SRIOV_1.
需要注意的是,控制节点上创建VNF使用的逻辑网络禁止绑定子网,因此标红加粗字段"subnets"后Value一栏必须为空.
如果Value一栏有子网值,可使用"neutronsubnet-deletesubname"(subname为子网名称)命令删除子网.
[root@ctrl1~]#novaquota-update--instances30--cores250--ram204800--server-groups2026975ace82564da3b24f873754a9cb94[root@ctrl1~]#novaquota-show|Quota|Limit||instances|30||cores|250||ram|204800||floating_ips|50||fixed_ips|-1||metadata_items|256||injected_files|2||injected_file_content_bytes|10240||injected_file_path_bytes|255||key_pairs|10||security_groups|15||security_group_rules|20||server_groups|20||server_group_members|10|5-37(2)查看和修改Openstack计算节点物理网络与网卡的映射关系登录到OpenStack计算节点的控制台命令行界面,查看和修改Openstack计算节点物理网络与网卡的映射关系.
A.
如果物理网卡为SRIOV类型网卡,通过vi编辑器打开/etc/neutron/plugins/ml2/ml2_conf_sriov.
ini文件,按[i]键进入编辑模式,找到physical_device_mappings选项,检查是否存在物理网络和SR-IOV网卡的映射关系.
如果不存在,手动配置映射关系.
此处以物理网络provider1与SR-IOV网卡ens1f0映射为例.
修改完成后,再通过vi编辑器打开/etc/nova/nova.
conf文件,按[i]键进入编辑模式,将物理网络与SR-IOV网卡的对应关系加入PCI白名单.
本例中以物理网络provider1与SR-IOV网卡ens1f0映射为例.
[root@ctrl1~]#neutronnet-shownetwork1-+|Field|Value|-+|admin_state_up|True||id|78c0a395-9e02-441a-8c33-9d6af1bdfb02||mtu|0||name|network1||provider:network_type|flat||provider:physical_network|provider1||provider:segmentation_id|||router:external|True||shared|True||status|ACTIVE||subnets|||tenant_id|26975ace82564da3b24f873754a9cb94|-+[root@compute1~]#vi/etc/neutron/plugins/ml2/ml2_conf_sriov.
iniphysical_device_mappings=provider1:ens1f0,phy_ens3f1:ens1f1[root@compute1~]#vi/etc/nova/nova.
confpci_passthrough_whitelist=[{"devname":"ens1f0","physical_network":"provider1"}]5-38最后,重启Nova和Sriov_Agent服务.
B.
如果物理网卡为普通网卡,通过vi编辑器打开/etc/neutron/plugins/linuxbridge/linuxbridge_conf.
ini文件,按[i]键进入编辑模式,找到physical_interface_mappings选项,检查是否存在物理网络和普通网卡的映射关系.
如果不存在,手动配置映射关系.
此处以物理网络provider1与普通网卡ens160映射为例.
修改完成后,重启Linuxbridge服务.
5.
查看Openstack控制节点和计算节点的Neutron代理服务状态是否正常登陆Openstack控制节点主机控制台的Linux命令行界面,通过"neutronagent-list命令"查看Openstack控制节点和计算节点的Neutron代理服务状态是否正常.
可通过回显信息中"admin_state_up"一栏值是否为"True"(即本例中的标红加粗字体部分)判断对应节点的Neuron代理服务是否正常,如果不为"True",则说明服务不正常,需要收集对应节点的Openstack相关日志.
6.
收集信息寻求技术支持如果按照上述方法检查完成后故障仍无法排除,请参考章节1.
2《收集故障信息》收集各类信息,拨打H3C热线400-810-0504寻求帮助.
[root@compute1~]#systemctlrestartopenstack-nova-compute.
service[root@compute1~]#systemctlrestartneutron-sriov-nic-agent.
servicenetwork1"}][root@compute1~]#vi/etc/neutron/plugins/linuxbridge/linuxbridge_conf.
iniphysical_interface_mappings=provider1:ens160,phy_ens192:ens192"}][root@compute1~]#systemctlrestartneutron-linuxbridge-agent.
service5-395.
1.
5故障诊断命令命令说明novahypervisor-list查看计算节点是否已和控制节点建立连接data查看系统时间date-s写入系统时间clock–w将系统时间同步到主机的硬件时间keystonetenant-list查看Openstack租户列表novaquota-show查看Openstack租户资源配额novaquota-update修改Openstack租户资源配额neutronnet-show查看Openstack逻辑网络systemctlrestartopenstack-nova-compute.
service重启OpenstackNova服务systemctlrestartneutron-sriov-nic-agent.
service重启SR-IOV服务neutronagent-list查看OpenstackNeutron代理服务状态5.
1.
6故障信息收集方法参考章节1.
2《收集故障信息》