Whenever the amcheck -c Full fails on the client the service in svcadm
will go to maintenance.

Here are the debug logs...
cat amandad.20071008155445018.debug
amandad: debug 1 pid 10963 ruid 30063 euid 30063: start at Mon Oct  8
15:54:45 2007
Could not open conf file "/etc/amanda/amanda-client.conf": No such file
or directory    --- not using a client file
amandad: time 0.002: amandad: invalid service  -- extremely odd
considering they are in inetd.conf and /etc/services.
amanda  dgram   udp     wait    amanda
/opt/amanda/client/libexec/amandad amandad

amanda    10080/udp    # amanda
amandaidx 10082/tcp    # amanda
amidxtape 10083/tcp    # Amanda





-----Original Message-----
From: Charlie Reitsma [mailto:reitsma AT denison DOT edu] 
Sent: Monday, October 08, 2007 3:51 PM
To: Krahn, Anderson
Subject: Re: client issues

"svcs | grep amanda" will probably show amanda in a maintenance state  
as the var/adm/messages says.

I always had difficulty getting amanda from inetd into smf. My notes  
say to run inetconv but I do recall having to use different command  
line switches to get amanda running and trying to modify the  
/var/svc/manifest/network/amanda-udp.xml by hand. Once it's running  
I've not had problems with it.

Quoting "Krahn, Anderson" <AKrahn AT gs1us DOT org>:

> Currently running Amanda 2.5.2p1 on solaris 10.
> The server is setup and I have 1 client working correctly. However,
when
> I try to add another client it is timing out on the check.
>
> Server:
> #/opt/amanda/server/sbin/amcheck -c Full
>
> Amanda Backup Client Hosts Check
> --------------------------------
> WARNING: prdapp03.transora.com: selfcheck request failed: timeout
> waiting for ACK
> Client check: 2 hosts checked in 30.197 seconds, 1 problem found
>
> Disk list
> /etc/amanda/Full]#vi disklist
> "disklist" 15 lines, 539 characters
> # doesn't work prdapp16.transora.com    /tmp always-full
> egvmgmt5001-prd.1sync.org / always-full      --- this server works.
> egvmgmt5001-prd.1sync.org /opt  always-full
> egvmgmt5001-prd.1sync.org /var always-full
> egvmgmt5001-prd.1sync.org /home always-full
>
> #prdapp01.transora.com / always-full
> #prdapp01.transora.com /opt always-full
> #prdapp01.transora.com /var always-full
> #prdapp01.transora.com /home always-full
>
> prdapp03.1sync.org / always-full
> prdapp03.1sync.org /opt always-full
> prdapp03.1sync.org /var always-full
> prdapp03.1sync.org /home always-full
>
> Client:
>
> /etc/inetd.conf
> amanda dgram udp wait amanda /opt/amanda/client/libexec/amandad
amandad
>
> inetconv was run and the service is online.
>
> netstat -a | grep amanda
>       *.amanda                            Idle
>
> error In /var/adm/messages
>
> Oct  8 12:58:16 prdapp03 inetd[426]: [ID 702911 daemon.error] Instance
> svc:/network/amanda/udp:default has exceeded its configured failure
> rate, transitioning to maintenance
> Oct  8 12:58:16 prdapp03 inetd[426]: [ID 702911 daemon.error] Instance
> svc:/network/amanda/udp:default has exceeded its configured failure
> rate, transitioning to maintenance
>
> cat .amandahosts
> prdapp16.transora.com amanda amdump
> prdapp16.1sync.com amanda amdump
> prdapp16 amanda amdump
>
> However my /tmp/client directory is empty.
>
> Amdump output:
>
> These dumps were to tape EGV002.
> The next tape Amanda expects to use is: EGV002.
>
> FAILURE AND STRANGE DUMP SUMMARY:
>   prdapp03.1sync.org  /      RESULTS MISSING
>   prdapp03.1sync.org  /opt   RESULTS MISSING
>   prdapp03.1sync.org  /var   RESULTS MISSING
>   prdapp03.1sync.org  /home  RESULTS MISSING
>   planner: ERROR Request to prdapp03.1sync.org failed: timeout waiting
> for ACK
>
> Strange is that I don't get any logs on the client side
> Host.debug
> Anderson
> ~
> ~
> ~
> ~
> ~
> ~
> ~
>



Charlie Reitsma
x6642