nsrexecd appears appears hung on HPUX clients

[SNunziata]

Having a problem backing up HPUX clients from a W2K Adv Svr. NetWorker server. Receiving the following errors:

(1) nsrexec: nsrexecd on <client> is unavailable.
Using rsh instead.
(2) connection timed out.
(3) nsrexec: remshd: Login Incorrect. Permission denied.

NetWorker server has NetWorker version 6.1.1. Build 238.
Doesn't matter what version of HP or NetWorker is on client machines. We use hosts files for name resolution. Hosts files are good to go. You can ping Networker server from client and vice versa. Starting a backup from the client using the nwrecover or save command works fine. When you start the backup from the NetWorker server is when you have a problem. We have bounced the daemons, reinstalled, tried different versions of NetWorker. Same poroblem. It seems that when the NetWorker server tries to invoke the nsrexecd from the client to start the save, savefs, it hangs then errors out.

We have been looking into a solution for this problem for two days and have tried alot of things. Any help would be appreciated.

/scott

 

[mkkeane]

Permission denied is a clear message that there is an authentication issue here. Since this is HP and it is trying to use rsh, perhaps this is an issue of user equivolence. configure a .rhost file and place the host name of the Win2k server in it. Also, you could try placing a username and password in the HP client setup (user account should be admin account of course). Either one of these things should work. Hope this helps.

Mkeane

 

[SNunziata]

Thanks for the reply. We tried editing the rhosts file to set the permissions of the root login on one of the clients failing in case the backups do not work using nsrexecd. Not the greatest way of doing it but none the less it would work in case of problems with the networker daemon.

Usually we have found out with Legato that every time you have a backup problem it is (in most cases) some sort of naming resolution issue between the legato server and client. In this case the etc/nsswitch.conf file had been modified on the clients that had a problem. We are still investigating why this had happened. An example of a correct nsswitch.conf file is:

hosts: files [NOTFOUND=continue UNAVAIL=continue] dns [NOTFOUND=return UNAVAIL=continue TRYAGAIN=return]

Apparently Networker has problems with the syntax of this file. If you are using the hosts file for naming resolution it is suggested that the hosts file is the first then dns. But the spaces are apparently an issue. If the spaces are not there, the backups will intermittantly fail. When you configure this on the HPUX machine using SAM, it does not automatically add the spaces between the &quot;hosts: file&quot; and the &quot;dns&quot; entries. This is a known problem with Legato and every new release of NetWorker that comes out has claimed to fix this.

Thanks for the reply again.

Have a good weekend.

/scott