Linux restore issue

[Bogdan_Visinesc]

Hi all,

I am trying to do something exotic.
I have backed up all filesystems from a SLES box. I have rebooted using the rescue mode from the installation CD on a another host (hardware the same). After I have partitioned the harddisk as it was on the original disk, I have managed to put up the ethernet network, and after that I have mounted via NFS the openv directory from another linux host. I have modified the bp.conf file in order to reflect the former client configuration.
Right now, my plan was to start the vnetd and bpcd daemons in standalone mode and after that to restore the files using a restore.rename file.
I have successfully started the vnetd daemon (/usr/openv/bin/vnetd -standalone), but whenever I have tried to start the bpcd daemon it gives me the following errors (from debug mode):

<2> ListenForConnection: vnet_vnetd_listen_and_register_service (bpcd) failed: 2
<16> bpcd main: ListenForConnection failed

Do you have any ideas on this issue?

thanks,
bogdan

[zippy]

Bogdan,

I'm thinking to myself; why are you doing this...

I will take a stab at it.

You stil have to have some sort of networking configuration done on the nfs mounted server, netbackup wont run just with the binaries

/usr/openv is nfs mounted, you forgot to add the /etc/services

bprd 13720/tcp # BPRD (VERITAS NetBackup)
bprd 13720/udp # BPRD (VERITAS NetBackup)
bpdbm 13721/tcp # BPDBM (VERITAS NetBackup)
bpdbm 13721/udp # BPDBM (VERITAS NetBackup)
bpjava-msvc 13722/tcp # BP Java MSVC Protocol
bpjava-msvc 13722/udp # BP Java MSVC Protocol
vnetd 13724/tcp # Veritas Network Utility
vnetd 13724/udp # Veritas Network Utility
bpcd 13782/tcp # VERITAS NetBackup
bpcd 13782/udp # VERITAS NetBackup
vopied 13783/tcp # VOPIED Protocol
vopied 13783/udp # VOPIED Protocol

And

xinetd and iptables, firewall and netwoking stuff, also there is a LIB patch to, search my ID for linux youll see it.

do you have a lot of time on your hands?

just thinking

JD

[Bogdan_Visinesc]

Hi James,

Thanks for your time.

I am doing this as an alternative to disaster recovery without using bare metal restore.

The file /etc/services has all the proper entries by default (this is SLES 9).

In rescue mode there are no firewall settings, nor iptables is run. Iptables is not even available. As I am trying to run the daemons in standalone mode I do not need the xinetd daemon to be running.

And be advised that this has worked on Itanium platform.

And yes, it appears as I am having quite a lot of free time. But the road to innovation is not paved at all.

I will appreciate any comments on my issue.

many thanks,
bogdan

[Stumpr2]

<2> ListenForConnection: vnet_vnetd_listen_and_register_service (bpcd) failed: 2
<16> bpcd main: ListenForConnection failed

James posted to check services
bpcd 13782/tcp # VERITAS NetBackup
bpcd 13782/udp # VERITAS NetBackup

are you listening? (netstat -a)

DOCUMENTATION: How to verify the required Veritas NetBackup (tm) 6.0 daemons / services are up and running on a client.
http://support.veritas.com/docs/282162

[Bogdan_Visinesc]

Problem solved.
The problem was that vnetd daemon was running, but bpcd daemon refused to start.
In UNIX world, vnetd creates a socket, on the filesystem where it resides. If that filesystem is on a NFS mount, even if you (even as root) have all the right permissions to create and delete files in that filesystem, you do not have the right to modify the file in order to create it as a socket. So, the resolution was to export the filesystem through NFS, using the NFS option of 'no_root_squash'.

Thanks for your time.

bogdan

[Bogdan_Visinesc]

Thank God for the existing 'strace' in a rescue linux environment.