Amanda-Users

Re: HOST DOWN error connecting to new client

2006-06-14 13:52:05
Subject: Re: HOST DOWN error connecting to new client
From: Kevin Till <kevin.till AT zmanda DOT com>
To: Matt Emmott <memmott AT jerkychew DOT com>
Date: Wed, 14 Jun 2006 10:49:02 -0700
Matt Emmott wrote:

-----Original Message-----
From: owner-amanda-users AT amanda DOT org [mailto:owner-amanda-users AT amanda 
DOT org]
On Behalf Of Matt Emmott
Sent: Wednesday, June 07, 2006 10:55 AM
To: 'George'
Cc: amanda-users AT amanda DOT org
Subject: RE: HOST DOWN error connecting to new client



-----Original Message-----
From: George [mailto:gkofoed AT rogers DOT com]
Sent: Tuesday, June 06, 2006 5:54 PM
To: Matt Emmott
Subject: Re: HOST DOWN error connecting to new client

Matt Emmott wrote:


Hi,

I just compiled and installed Amanda 2.5.0p2 as a client on a Red Hat
Enterprise ES 4 machine. Our server (which has been running for years)
is a Red Hat 8 machine running Amanda 2.4.2p2. I've modified the
disklist file in the archive section of our server to only back up the
new server, named owl, via the following line:

owl /home/path/to/files user-tar

However, when I try to run /usr/sbin/amcheck Archive, I get the
following error:

===============================================================

Amanda Tape Server Host Check

-----------------------------

Holding disk /usr/amandatmp: 10851752 KB disk space available, that's
plenty

ERROR: label DailySet103 doesn't match labelstr "^archive[0-9][0-9]*$"

(expecting tape archive09 or a new tape)

NOTE: skipping tape-writable test

NOTE: info dir /var/lib/amanda/Archive/curinfo/owl: does not exist

NOTE: index dir /var/lib/amanda/Archive/index/owl: does not exist

Server check took 18.177 seconds

Amanda Backup Client Hosts Check

--------------------------------

WARNING: owl: selfcheck request timed out. Host down?

Client check: 1 host checked in 29.998 seconds, 1 problem found

===============================================================

This was just a test run, so the bad tape error is ok, and I assume
the info and index errors are ok, since a backup has never been run.
The problem is with the Host Down error. I've enabled Amanda in
xinetd.d, and restarted xinetd. a chkconfig --list tells me that
Amanda is up and running. There is no firewall on the server, and
hosts.deny is empty.

What else can I look at to figure out why this isn't connecting?

Thanks,

Matt


==================

Hi Matt;

A couple of things to check:

- check the amanda tmp (/tmp/amanda or /var/log/amanda) directory on the
client for any log files or diagnostics (selfcheck.<date>.debug or
amandad.<date>.debug). If there are no such files in your amanda temp
directory, then your client is not "hearing" any comms from the server;


- make sure that the xinetd setup for amanda is not disabled (default).
My RHEL clients have the following entry
service amanda {
disable = no
socket_type = dgram
protocol = udp
wait = yes
user = amanda
group = disk
server = /usr/lib/amanda/amandad
}

- within the amanda xinetd config, the "server" statement must include
the correct path to the "amandad" binary. Emphasis on "correct", I've
been bitten by this one more than once.

- ensure there's an entry in /etc/services for amandad. My clients have
the following (I suspect this is your problem):

amanda 10080/tcp # amanda backup services
amanda 10080/udp # amanda backup services

If you make any changes to the xinetd config or /etc/services, ensure
you restart xinetd. Hope this helps!

George


Hi George,

I have 2 .debug files in /tmp/amanda. Here's the output of the newest one:

======================================
amandad: debug 1 pid 27781 ruid 501 euid 501: start at Tue Jun  6 17:04:31
2006
amandad: version 2.5.0p2
amandad: build: VERSION="Amanda-2.5.0p2"
amandad:        BUILT_DATE="Thu May 11 15:53:13 EDT 2006"
amandad:        BUILT_MACH="Linux owl 2.6.9-5.ELsmp #1 SMP Wed Jan 5
19:30:39 EST 2005 i686 i686 i386 GNU/Linux"
amandad:        CC="gcc"
amandad:        CONFIGURE_COMMAND="'./configure' '--with-user=amanda'
'--with-group=amanda'"
amandad: paths: bindir="/usr/local/bin" sbindir="/usr/local/sbin"
amandad:        libexecdir="/usr/local/libexec" mandir="/usr/local/man"
amandad:        AMANDA_TMPDIR="/tmp/amanda" AMANDA_DBGDIR="/tmp/amanda"
amandad:        CONFIG_DIR="/usr/local/etc/amanda" DEV_PREFIX="/dev/"
amandad:        RDEV_PREFIX="/dev/" DUMP="/sbin/dump"
amandad:        RESTORE="/sbin/restore" VDUMP=UNDEF VRESTORE=UNDEF
amandad:        XFSDUMP=UNDEF XFSRESTORE=UNDEF VXDUMP=UNDEF
VXRESTORE=UNDEF
amandad:        SAMBA_CLIENT="/usr/bin/smbclient" GNUTAR="/bin/gtar"
amandad:        COMPRESS_PATH="/bin/gzip" UNCOMPRESS_PATH="/bin/gzip"
amandad:        LPRCMD="/usr/bin/lpr" MAILER="/usr/bin/Mail"
amandad:        listed_incr_dir="/usr/local/var/amanda/gnutar-lists"
amandad: defs:  DEFAULT_SERVER="owl" DEFAULT_CONFIG="DailySet1"
amandad:        DEFAULT_TAPE_SERVER="owl" DEFAULT_TAPE_DEVICE="null:"
amandad:        HAVE_MMAP HAVE_SYSVSHM LOCKING=POSIX_FCNTL SETPGRP_VOID
amandad:        DEBUG_CODE AMANDA_DEBUG_DAYS=4 BSD_SECURITY RSH_SECURITY
amandad:        USE_AMANDAHOSTS CLIENT_LOGIN="amanda" FORCE_USERID
HAVE_GZIP
amandad:        COMPRESS_SUFFIX=".gz" COMPRESS_FAST_OPT="--fast"
amandad:        COMPRESS_BEST_OPT="--best" UNCOMPRESS_OPT="-dc"
amandad: time 29.954: pid 27781 finish time Tue Jun  6 17:05:01 2006
======================================

My xinetd setup for Amanda appears to be correct, and the path specified
in
the server section does exist.

/etc/services has several entries for Amanda. They are:

amanda          10080/tcp                       # amanda backup services
amanda          10080/udp                       # amanda backup services
kamanda         10081/tcp                       # amanda backup services
(Kerberos)
kamanda         10081/udp                       # amanda backup services
(Kerberos)
amandaidx       10082/tcp                       # amanda backup services
amidxtape       10083/tcp                       # amanda backup services

Any other ideas?

Thanks!
Matt



I've rebooted both servers to rule out any network wackiness, and I'm still
having the same problem. A chkconfig shows that the Amanda service is in
fact running under xinetd. I was wondering if there's any other way to
connect directly to the machine to see if it is in fact answering requests
on those ports? Would a simple NMAP scan work? Can I telnet to the ports and
get some kind of response?


Since amanda*.debug on the client got created, it seems to me the server did the initial contact to the client successfully. Just selfcheck didn't get executed.
Do you have the following file on the client:
/usr/local/libexec/selfcheck
/usr/local/libexec/noop
/usr/local/libexec/sendsize

Yes, "nmap -sU -sV host" should report 10080/udp is open on the client.




--
Thank you!
Kevin Till

Amanda documentation: http://wiki.zmanda.com
Amanda forums:        http://forums.zmanda.com