10-02-2008 01:08 PM
Hi,
I have just installed the netbackup client on some new Linux servers.
Netbackup 5.1 MP4
Oracle Enterprise Linux 5.1
When I try to do the Host Properties it timesout with "invalid error number"
The bprd log shows -
15:55:10.609 [9913] <2> bprd: socket fd from accept() is 7
15:55:10.611 [9913] <2> listen_loop: request complete
15:55:10.625 [10544] <2> logconnections: BPRD ACCEPT FROM 10.10.0.100.2771 TO 204.76.244.5.13720
15:55:10.628 [10544] <2> process_request: setsockopt SO_LINGER on 7 succeeded.
15:55:10.633 [10544] <2> connected_peer: Connection from host swf11848.ad.swfwmd.net, 10.10.0.100, on non-reserved port 2771
15:55:10.712 [10544] <2> process_request: command C_READ_HOST_CONFIG (73) received
15:55:10.713 [10544] <2> get_type_of_client_port: db_getCLIENT() failed: no entity was found (227)
15:55:10.714 [10544] <2> nb_getsockconnected: Connect to bkvlrac01.ad.swfwmd.net on port 599
15:55:10.714 [10544] <2> logconnections: BPCD CONNECT FROM 204.76.244.5.599 TO 10.10.24.30.13782
15:55:22.530 [10507] <2> bpcr_connect: bpcr_connect timeout during select after 60 seconds on port 864
15:55:22.626 [9913] <2> listen_loop: select() interrupted
15:56:10.721 [10544] <2> bpcr_connect: bpcr_connect timeout during select after 60 seconds on port 526
15:56:10.809 [9913] <2> listen_loop: select() interrupted
I can ping from the master-server to the client and vice versa.
I can nslookup from the master-server by both hostname and ip address - and vice versa.
I issued the bpclntcmd on both the master-server and client as well as another linux server that is working fine and -hn for hostname, -ip for ipaddress, and -self work. When I issue the -pn switch on the client it only returns a command prompt and the bprd log shows -
13:33:17.136 [9913] <2> bprd: socket fd from accept() is 7
13:33:17.157 [9913] <2> listen_loop: request complete
13:33:17.178 [22406] <2> logconnections: BPRD ACCEPT FROM 10.10.24.30.13301 TO 204.76.244.5.13720
13:33:17.189 [22406] <2> process_request: setsockopt SO_LINGER on 7 succeeded.
13:33:17.198 [22406] <2> connected_peer: Connection from host bkvlrac01.ad.swfwmd.net, 10.10.24.30, on non-reserved port 13301
13:33:17.214 [22406] <2> db_valid_master_server: bkvlrac01.ad.swfwmd.net is not a valid server
13:33:17.214 [22406] <2> get_long: (2) premature end of file (byte 1)
13:33:17.260 [22406] <32> process_request: get_string() failed - premature end of file encountered (5)
13:33:17.264 [22406] <32> process_request: could not process request from bkvlrac01.ad.swfwmd.net
13:33:17.345 [9913] <2> listen_loop: select() interrupted
bp.conf looks fine on the client.
I would appreciated any help with this issue!
Thanks.
Solved! Go to Solution.
10-03-2008 09:16 AM
Check xinetd for proper configuration.
Try using "bpclntcmd" to test basic connectivity from NetBackup's perspective.
use "netstat -a" to check that the appropriate ports are open.
I also believe that Linux has it's own build in firewall that you may want to check to see who it's configured. (I''m not familiar with it though).
10-02-2008 01:50 PM
as this is a new client and the first time you are trying to connect to it from the master server you need to do the first steps of on communication.
check the dns entries for the new client.
does it have both a forward and reverse.
can you ping the client from the master both by name and ip
can you ping the master from the client both by name and ip
check to see if you have any ports closed that nb needs.
10-03-2008 05:31 AM
Thanks for you quick response!
I have checked the dns and all is working. I can nslookup from the master-server to the client by hostname and ip. Also, nslookup from the client to the master-server by hostname and ip. Ping also works both directions by hostname and ip address.
I do not have any ports closed (that I know of) - is there a command I can issue to verify that. Also, the bprd log does show that a connection is made -
connected_peer: Connection from host bkvlrac01.ad.swfwmd.net, 10.10.24.30, on non-reserved port 13301
As I said, this is on Oracle Enterprise Linux (Red Hat based) 5.1 kernel 2.6.18-53.el5. I have three servers with this os and all three have the same issue.
Thanks!
10-03-2008 09:16 AM
Check xinetd for proper configuration.
Try using "bpclntcmd" to test basic connectivity from NetBackup's perspective.
use "netstat -a" to check that the appropriate ports are open.
I also believe that Linux has it's own build in firewall that you may want to check to see who it's configured. (I''m not familiar with it though).
10-03-2008 11:09 AM
Thanks for the quick response. I have used the "bpclntcmd" and it works for the -hn and -ip switches, not for the -pn.
Xinetd is running and my xinetd.conf is the default - is the a specific entry needed?
In the linux desktop 'Service Configuration' I see "bpcd" and "bpjava-msvc' checked.
The built in firewall is disabled.
Thanks!
10-03-2008 01:55 PM
Test the bpcd port on the master:
From the master: telnet localhost 13782
From the client: telnet <master> 13782
It should connect, wait for you to hit ENTER and then drop you off. Does that happen or do you get an error?
10-03-2008 02:03 PM
With the mention of xinetd.conf and some more googling ;) I was able to get it to work.
Added the following to xinetd.conf -
service bpcd
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/openv/netbackup/bin/bpcd
# server_args = bpcd
#only_from = ........removed...........
}
service vopied
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/openv/netbackup/bin/vopied
# server_args = vopied
#only_from = ........removed...........
}
service bpjava-msvc
{
socket_type = stream
protocol = tcp
wait = no
user = root
server = /usr/openv/netbackup/bin/bpjava-msvc
server_args = -transient
#only_from = ........removed...........
}
Thanks to all for the help!