Unix agent password failure

[mleggatt]

I have a Windows NT server with the ARCSERVE manager installed backing up a Solaris 5.7 Unix box. Until recently this worked o.k. The power supply to the box failed and the machine had to be rebooted. On rebooting it appears that I now cannot get past the USERNAME and PASSWORD screen when trying to browse down the tree on the Backup manager screen. I have tried re-installing the agent and applying the latest service pack (SP1 Build 884) but when I browse down the tree in the Backup manager window to the Unix box in question (under Unix agents) the logon dialogue comes up and I enter the root user name and password but get a "Login to agent" failed message. I know that this password and username is correct as I can Telnet to the box using it. I found a log file within /USR/CYEagent/logs which seems to show what the error is but I have no idea what is means (see below).

6/04 17:38:04(1515) - (getRequest) enter ...
06/04 17:38:04(1515) - (getRequest) blkOffset=680
06/04 17:38:04(1515) - (getRequest) objType=0, opMode=14
06/04 17:38:04(1515) - (agentLogin) loginID=[], paswd=[*]
06/04 17:38:04(1515) - (hostequiv) peerAddr=[192.168.xxx.xxx]
06/04 17:38:04(1515) - (ckpasswd) Failed in getpwnam(), Socket operation on non-socket
06/04 17:38:04(1515) - -- Object [0] [/opt/uagent] ------------------
06/04 17:38:04(1515) - agent info: [Uagent] 5.0.0.2 0x0

I have looked on Google e.t.c for this but found nothing. Has anyone else experienced this or resolved the issue?? any pointers would be appreciated.

Thanks
Martin

 

[SjrH]

Try deleteing the Solaris box from the agents list and re-discovering it again.

Rgds

SjrH - Data Security Storage & Availability Specialist

http://www.sjrh.co.uk

 

[mleggatt]

Thanks for the advice, but after several install and lots of head scratching I found the cause of the problem. It was due to the entries that I added to /etc/services and /etc/inet/inetd.conf to automatically start the uagentd daemon. I added these entries so that when the Unix boxes were restarted and the ARCSERVE server tried to connect to the Unix box the uagentd daemon would start automatically as suggested in the Computer Associates support site. I am not sure if this support article is out of date and the uagentd agents have been updated to be started as a daemon in the init(rc) scripts but when I looked in the syslog on the solaris boxes I found an error saying that the uagentd was looping and had been terminated (about 76 loops). I think what was happening was the uagentd was running, inetd then recieved multiple reuqests from the ARCSERVE server with each request spawing another uagentd until a limit was reached which then caused the uagentd processes to be terminated. This meant that when trying to access the uagent from the ARCSERVE server it seemed as thought it could not connect.

I am currently investigating whether the uagentd install scripts place an entry in the rc scripts so that it starts automatically and if not how I do this myself.

thanks again for the advice, I thought I had better post this so that anyone else with this issue can resolve it.

Just a note it seems to me that the ARCSERVE software is very buggy and the support is patchy, incorrect or very difficult to find. This forum has more replies per post than ARCSERVE's own forums!!