When I try to launch a backup from my client, nothing happened, with user backup or immediate backup. But when I try to directly launch the policy, it works.
My policy contains 2 User Backup schedules and one Full Backup schedule.
As I said, when I launch manually the policy, backup works.
But when I try to launch bpbackup command from client, nothing is happening.
I try the following commands from the client
bpbackup -p PL-UNIX-USERDATA-OPS-ME -s 1month /tmp
bpbackup -i -p PL-UNIX-USERDATA-OPS-ME -s test
From the master, the command bpbackup -i -p PL-UNIX-USERDATA-OPS-ME -s test works.
I can see that in the bprd logs, but I don't know what does it means
11:28:57.869  <2> vnet_pbxAcceptSocket: Accepted sock from 10.94.19.30:54019
11:28:57.869  <2> bprd: socket fd from accept() is 9
11:28:57.870  <2> listen_loop: request complete
11:28:57.870  <2> logconnections: BPRD ACCEPT FROM 10.94.19.30.54019 TO 10.30.34.129.1556 fd = 9
11:28:57.870  <2> process_request: setsockopt SO_LINGER on 9 succeeded.
11:28:57.873  <2> connected_peer: Connection from host pars3popsbck001.eu.net.intra, 10.94.19.30, on non-reserved port 54019
11:28:57.877  <2> process_request: magic  cmd  ...
11:28:57.877  <2> process_request: ...request ignored...continuing...
My client (pars3popsbck001) is in NBU 8, and my master (eqds3pnbmam001) is in NBU 188.8.131.52.
I've attached logs of bpbackup from client
That is not what I said.
NetBackup 8 Client is not at all supported with lower version Master/media server.
(See Release Notes for each NBU version)
So, you are extremely lucky that scheduled backups are working.
You need to upgrade your NBU master/media server(s).
Level 3 or higher bprd log may tell us the reason for "...request ignored.. "
The reason why I asked for policy config is because the client sends the configured client name with the backup request.
This does not happen with a scheduled backup.
We see in client log that the client sent to bprd is pars3popsbck001.
NBU master resolves the IP address to pars3popsbck001.eu.net.intra.
This could be problem if no aliases exist in DNS or hosts file, but then I would expect an error.
I would also like to see the schedule config in the policy.
If /tmp is on tmps filesystem, it will be automatically excluded.
I've tested with /tmp only because it's a small FS, to see if job arrive in activity monitor.
But I've tested with a master in 8.0 and backup is working from the client, so it seems that it's due to compatibility problem. And I don't want to backup 8 client with a lower master, to avoid all possible problems.
I will find another way to backup my datas as we can't upgrade for the moment.