分类： 服务器与存储
2011-01-01 23:29:58
1. 问题描述
在LINUX 客户端安装完netbackup client和netbackup for db等netbackup客户端后，在netbackup服务器端定义该linux客户端的备份策略，完成备份策略的定义后,需要定义客户端的一些属性;NetBackup Management-->Host Properties-->Clients-->双击右边窗口的目标host进行客户端的属性设置时,报cannot connect on socket,错误的代码为:status code=25。
2. 问题分析
2.1 hostname 定义检查
通过检查，发现不是hostname问题，在netbackup服务器和客户端的hosts文件都有IP与主机名对应的正确定义。
2.2 检查客户端是否安装了xinetd服务
检查linux客户端，发现没有安装xinetd服务，担心是因为这个问题导致了code＝25的错误。
#rpm -aq|grep xinetd
没有任何输出内容，说明xinetd包没有安装。
2.3 在服务端测试连接客户端的netbackup client应该打开的端口
使用telnet测试客户端的端口是否打开
c:>telnet testdb 13782
经测试，发现服务端服务连接客户端的13782端口，初步判断，客户端安装以后，一些相关的服务没有启动正确。
问题最终分析：
linux客户端安装netbackup一定需要先安装xinetd服务，netbackup客户端最核心的服务bpcd是xinetd的一个子服务，如果不安装xinetd服务，则bpcd无法正确启动，bpcd不启动，客户端13782端口就无法启动。
导致本次报code＝25错误的原因有两个
A. linux客户端没有安装xinetd服务
B. linux客户端没有配置好bpcd服务
3. 问题解决
3.1下载xinetd包,并进行xinetd包的安装.
#uname -a
确定操作系统的版本,然后去网上下载对应的xinetd包.
#rpm -Uvh /tmp/xinetd-2.3.14-10.el5.i386.rpm
安装xinetd包
#rpm -aq|grep xinetd
xinetd-2.3.14-10.el5
3.2 重新安装netbackup客户端软件
安装完xinetd服务后,检查xinetd的配置文件/etc/xinetd.conf,发现没有netbackup的服务bpcd的定义;然后检查/etc/xinetd.d目录内容,也没有bpcd的内容;
#cd /etc/xinetd.d
#ls
chargen-stream daytime-stream echo-dgram ekrb5-telnet krb5-telnet tcpmux-server vnetd
……
重新安装netbackup client和netbackup for db等客户端软件,然后再检查/etc/xinetd.d目录,发现有了bpcd服务的定义了.
#ls /etc/xinetd.d
bpcd chargen-stream daytime-stream echo-dgram ekrb5-telnet krb5-telnet tcpmux-server vnetd
……
#more /etc/xinetd.d/bpcd
service bpcd
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/openv/netbackup/bin/bpcd
disable = no
}
注意: 有时候安装完xinetd服务后,再安装netbackup client,netbackup for db等客户端软件后,在/etc/xinetd.d/目录下还是没有bpcd文件(笔者曾经遇到这个问题,并重新安装了几次都是这样).我们可以通过手工创建该文件来解决这个问题.
#vi /etc/xinetd.d/bpcd
#文件内容如下:
service bpcd
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/openv/netbackup/bin/bpcd
disable = no
}
3.3 重新启动xinetd服务
#service xinetd restart
Stopping xinetd: [ OK ]
Starting xinetd: [ OK ]
[root@yxcarddata xinetd.d]# service xinetd status
xinetd (pid 32397) is running...
3.4检查bpcd是否启动
#netstat -a|grep bpcd
tcp 0 0 *:bpcd *:* LISTEN
说明bpcd 已经启动了,处于listen状态.
3.5 检查netbackup的bpcd的端口
# more /etc/services|grep bpcd
bpcd 13782/tcp # VERITAS NetBackup
bpcd 13782/udp # VERITAS NetBackup
如果客户端与服务器之间存在防火墙,则需要开放服务器到客户端之间的13782(tcp,udp)的端口访问权限,需要开放客户端到服务器端的13720(13724)的端口访问.具体的端口可以联系网管员,检测防火墙的拦截信息,然后根据实际情况进行端口访问开放.
最后去netbackup服务器上,点击客户端,不再报status code=25的代码错误了,可以对客户端的属性修改和维护了.
附件:
关于netbackup的code=25的错误描述,netbackup 给出了问题描述与问题解决,摘录如下:
Problem:
A process that timed out while connecting to another process for a particular operation. This problem can occur in the following situation: when a process tries to connect to the NetBackup request daemon (bprd) or database manager daemon (bpdbm) and the daemon is not running. (On Windows, these daemons are the NetBackup Request Manager and NetBackup Database Manager services.)
It can also occur in the following situations: the network or server is heavily loaded and has slow response time or an evaluation license key for NetBackup expired. However, the most common cause of this error is a host name resolution problem.
The following are other possible causes of this error caused either by network connectivity issues or if a required process such as pbx_exchange is not running.
* nbjm is unable to connect to bpcd on the media server
* nbpem is unable to connect to nbproxy
* bptm on the media server is unable to connect to nbjm on the master server.
* You cannot perform an immediate backup operation.
These errors are
Troubleshoot:
Do the following, as appropriate:
* Verify that bpcompatd, vnetd, and Private Branch Exchange (PBX) are running. Information on how to start PBX is available.
See "Resolving PBX problems" in the Troubleshooting Guide.
*
If necessary, stop and restart NetBackup.
* On UNIX and Linux systems, enter the following:
/usr/openv/netbackup/bin/bp.kill_all
/usr/openv/netbackup/bin/bp.start_all
* On Windows systems, enter the following:
install_pathNetBackupbinbpdown
install_pathNetBackupbinbpup
* The following information applies only to a UNIX or Linux NetBackup master server:
Verify that the bprd and the bpdbm processes are running. If these processes are not running, start them. On a Windows master server, verify that the NetBackup Request Manager and NetBackup Database Manager services are running. If these services are not running, start them.
If these processes or services are running, examine the All Log Entries report for the time of the failure to determine where the failure occurred.
Do one of the following:
* If you cannot view the report or you get a cannot connect on socket error when you try to view it, do the following: verify again that the NetBackup Database Manager service or daemon is running. Then, create a debug log directory for bpdbm, retry the operation, and check the resulting debug log.
* If you can view the report and have not found an entry that is related to this problem: create debug log directories for the related processes that were running when the error first appeared. (This process frequently is bpbrm.) Then, retry the operation and check the resulting debug logs.
* Verify that the server list specifies the correct master server.
* The following information applies only to Windows systems:
The master server is designated in the Server to use for backups and restores drop-down in the Specify NetBackup Machines and Policy Type dialog box. To display this dialog box, start the Backup, Archive, and Restore interface and click Specify NetBackup Machines and Policy Type on the File menu.
* The following information applies only to UNIX, Linux, and Macintosh systems:
The master server is the first SERVER entry in the bp.conf file.
* On NetWare target clients: the master server name is the first SERVER entry in the bp.ini file.
* Make sure all recommended NetBackup patches were installed. Check the Symantec support Web site for current patch information.
Go to .
Then select "NetBackup" followed by "files and updates."
* If failure occurs when you run a user-directed backup from a client, make sure a user-directed backup schedule exists at the master server.
* With NetBackup database extensions: make sure that the applicable database product has the correct permissions allowing NetBackup to write to the progress log on the client.
* The following information applies only to UNIX and Linux systems:
If bpdbm has quit when the shutdown script runs on a media server, carefully read the K77netbackup script. It contains details on how to prevent this problem. The script is in /usr/openv/netbackup/bin/goodies.
If you change the server list on a master server, stop and restart the following: the NetBackup database manager and request daemons (UNIX and Linux) or the NetBackup Database Manager and NetBackup Request Manager services (Windows).
* Check the services file.
* The following information applies only to UNIX and Linux systems:
Verify that the /etc/services file (and NIS services if NIS is used) has entries for the NetBackup services: bpcd, bpdbm, and bprd.
On Windows, verify that the %SystemRoot%system32driversetcservices file has the correct entries for bpcd, bpdbm, and bprd.
Also, verify that the following numbers match the settings in the services file: the NetBackup Client Service Port and the NetBackup Request Service Port on the Network tab in the NetBackup Client Properties dialog box. To display this dialog box, start the Backup, Archive, and Restore interface and click NetBackup Client Properties on the File menu. The values on the Network tab are written to the services file when the NetBackup Client service starts.
See "Verifying host names and services entries" in the Troubleshooting Guide.
* On Sun Solaris, verify that all operating system patches are installed
See the Operating Notes section of the NetBackup Release Notes.
* On Windows, verify that the recommended service packs are installed.
* When the base NetBackup license key expires, daemons (such as bprd and bpdbm) terminate on the NetBackup server. If these daemons are not running, you are likely to encounter status code 25 errors in the Administration console. Install a valid base NetBackup license key, restart the daemons, and restart the console.
* For NetBackup Snapshot Client, the following applies: When many devices are configured on a media server, it may take a long time for the bptpcinfo command to generate the 3pc. file. When the backup is run for the first time, the backup may fail with status 25. Make sure that the /usr/openv/volmgr/database/3pc.conf file exists. If it does, rerun the backup. If the backup fails again, run the bptpcinfo manually to generate the 3pc. file, then try the backup again.
