NB7.6.0.3 restore to different location issue

[Slawomir]

Hi.

Evnironment: MasterServer: Linux SLES11

RemoteConsoleServer: Windows 2008R2 (virtual machine) - is use for manage environment, backup polices, etc

MediaServers & Storage: 2x NetBacakup Appliance 5230 

NetBackup ver 7.6.0.3 is installd on all servers & clients.

Problem: the restore is failed on the same host but for different location.

8/26/2015 8:29:49 AM - Info bpbrm(pid=11935) "ClientName" is the host to restore to      
8/26/2015 8:29:49 AM - Info bpbrm(pid=11935) reading file list for client        
8/26/2015 8:29:50 AM - Error bpbrm(pid=11935) Request from host "MediaServer_Name" to host "RemoteConsoleServer_Name" is not allowed access. Host "MediaSerever_Name" is not an authorized server for host "MasterServer_Name".
8/26/2015 8:29:50 AM - Error bpbrm(pid=11935) bpcd on "MasterServer_Name" exited with status 59: access to the client was not allowed
8/26/2015 8:29:50 AM - Error bpbrm(pid=11935) cannot put rename file on "ClientName" 
8/26/2015 8:29:50 AM - Info tar(pid=0) done. status: 59: access to the client was not allowed   
8/26/2015 8:29:50 AM - Error bpbrm(pid=11935) client restore EXIT STATUS 59: access to the client was not allowed
8/26/2015 8:41:35 AM - begin Restore
8/26/2015 8:41:35 AM - restoring image "ClientName"_1440178202
8/26/2015 8:41:36 AM - Info bprd(pid=1693) Restoring from copy 1 of image created Fri Aug 21 19:30:02 2015 from policy "PolicyName"
8/26/2015 8:41:36 AM - requesting resource @aaaab
8/26/2015 8:41:36 AM - granted resource MediaID=@aaaab;DiskVolume=PureDiskVolume;DiskPool="DiskPool_Name";Path=PureDiskVolume;StorageServer="StorageServer_Name";MediaServer="MediaServer_Name"
8/26/2015 8:41:39 AM - restored image "ClientName"_1440178202 - (access to the client was not allowed(59)); restore time 0:00:04
8/26/2015 8:41:39 AM - Warning bprd(pid=1693) Restore must be resumed prior to first image expiration on Mon Sep 21 19:30:02 2015
8/26/2015 8:41:39 AM - end Restore; elapsed time: 0:00:04
MS-Windows policy restore error(2808)

What I've verified:

Host Propeties:

Master Server have all media & remote console on Additional Servers list

Media Servers have all media & remote console on Additional Servers list

Client have all media & remote console on Additional Servers list

Restore to the source place (original) is succesfully completed(0), but the same error is existing in logs

Problem existing for Windows & Linux clients. I don't check NDMP, SQL or VM yet.

[Marianne]

Do you have bprd log folder on the Linux master?
If so, please copy bprd log file for 26 Aug to bprd.txt and upload as File attachment.

If not, please create the folder under /usr/openv/netbackup/logs and restart NBU.
Retry the restore.
After failure, copy bprd log to .txt and upload as File attachment.

We may need more logs, but initial job setup is what we need to look at first.

[Slawomir]

attached log file

[Marianne]

I have difficulty trying to map hostnames seen in bprd log to the replaced names in your opening post.

If you don't mind, I will quote partial names and you can help me understand where they fit in the picture.

12:49:04.571 [39327] <2> process_request:   browse_clnt: app013
12:49:04.571 [39327] <2> process_request:   requesting_clnt: BAK0100
12:49:04.571 [39327] <2> process_request:   destination_clnt: app013
12:49:04.571 [39327] <2> process_request:   clnt_bp_conf_name: bak0200
12:49:04.571 [39327] <2> process_request:   peername: bak0100

What is the role of bak0100 and bak0200?

bak0200 seems to be the master, right?

It seems that restore is initiated from the GUI on bak0100, but for some reason, this machine's configured Client_Name seems to bak0200.

Media server used for the restore seems to be ba0201, right?

From what I can see is that bpbrm on media server ba0201 receives status 59 back from the client.

It seems that client app013 may not have a SERVER entry for media server ba0201.

The following logs will confirm:
bpbrm on media server ba0201 
bpcd on client app013

[Slawomir]

Hi Marianne.

What is the role of bak0100 and bak0200? - bak0100 = GUI administration console (Windows), bak0200 = Master Server (Linux).

I use bak0100 as cosole for managing NetBackup environment.

Media server = ba0201

Logs are attached. bpcd from client, bpbrm form media.

[Slawomir]

log from client

[Marianne]

Before I look at logs - please check the Client_Name entry in the registry of bak0100 - 

HKey_Local_Machine - Software - Veritas - Netbackup - CurrentVersion - Config

It must be the local hostname - not master server name.

It seems that this entry is 'confusing' the initial setup of the restore process.

You may want to check this entry for all roleplayers - master, media, Windows Admin console as well as destination client.
Registry on Windows machines and bp.conf on Unix/Linux machines.
All of them need to have Server entries as follows:

Server = bak0200 
Server = bak0100 
Server = ba0201

Each of them must have Client_Name the same as local hostname.

While you are double-checking all of this, I will have a look at logs. 
 

PS: 
If at all possible, please upload logs as .txt - I can only open zipfiles while working on my laptop.
Most of my Connect activity is in my private time using a mobile device where I cannot open zip files.

[Marianne · Accepted solution]

It seems that the incorrect Client_Name on the Windows Admin Console is causing the media server to try and connect to bak0100 as if it is bak0200:

Request from host ba0201 (172....19) to host BAK0100 (172....47) is not allowed access.
Host ba0201 (172....19) is not an authorized server for host bak0200. 

So, it is bpcd on bak0100 failing with status 59.

Please check and fix Client_Name entries as per my previous post.