RESULTS MISSING

[Rob Flory]

Hi,

As of yesterday my AMANDA backups stopped running, and simply send me a
mail report listings all my hosts with "RESULTS MISSING". I've reviewed
sendsize.debug on each of the hosts and it looks like everything was
working normal. I'm using the fully qualified hostname for the machine
local to AMANDA. What other reasons are there for AMANDA to give up and
report "RESULTS MISSING" for _EVERY_ host? I'm running Amanda version
2.4.1p1 on each host.

Thank You,
Rob

Re: timed out

[Rob Flory]

amandad: debug 1 pid 14470 ruid 50 euid 50 start time Mon Apr  9
19:54:21 2001
amandad: pid 14470 finish time Mon Apr  9 22:07:40 2001

The strange thing is that this client has been working fine for months
in the past and is now suddenely failing daily.

Suggestions?

Thanks,
Rob

Alexandre Oliva wrote:
> 
> On Apr  6, 2001, Rob Flory <[EMAIL PROTECTED]> wrote:
> 
> > amandad: waiting for ack: Connection refused, retrying
> 
> > does this mean corp could not make a connection back to the amanda
> > server?
> 
> Yep.  It probably means the server already timed it out.  What is the
> time difference between the first and the last line in amandad.debug?
>

Re: timed out

[Rob Flory]

amanda-2.4.1p1


from corp:
amandad: waiting for ack: Connection refused, retrying
amandad: waiting for ack: Connection refused, retrying
amandad: waiting for ack: Connection refused, retrying
amandad: waiting for ack: Connection refused, retrying
amandad: waiting for ack: Connection refused, giving up!
amandad: pid 20307 finish time Thu Apr  5 22:16:55 2001

does this mean corp could not make a connection back to the amanda
server?
or that the amanda server could not make a connection to corp?

corp is a real hostname, no other hostnames point to that machine in
disklist

Thanks,
Rob


"John R. Jackson" wrote:
> 
> >  corp   hdc3 lev 0 FAILED [Request to corp timed out.]
> >...
> >this error happens every night, i have already verified that:
> >...
> 
> What version of Amanda?
> 
> Are the /tmp/amanda/*.debug files being updated from the amdump run?
> Make sure you don't run amcheck after amdump before looking as it will
> clobber the amandad*debug file (fixed in 2.4.2p2).
> 
> Is "corp" a real host name?  Do any other host names listed in disklist
> point to the same real machine as "corp"?
> 
> >Rob
> 
> John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]

timed out

[Rob Flory]

I'm not sure why this happens:

FAILURE AND STRANGE DUMP SUMMARY:
  corp   hdc3 lev 0 FAILED [Request to corp timed out.]
  corp   hda3 lev 0 FAILED [Request to corp timed out.]
  corp   hda1 lev 0 FAILED [Request to corp timed out.]

this error happens every night, i have already verified that:

- the client (corp) is properly configured (correct entries are in
/etc/inetd.conf) and is listening and responding on the proper ports
- the server does not find this problem on client (corp) when i run
./amcheck
- the server and user (amanda) are set up in .amandahosts and .rhosts on
the client (corp)
- there is no firewall preventing direct communication between client
and server.

i've run amandad on client (corp) manually and it runs without err.


Any suggestions?
thanks in advance,

Rob