Amanda-Users

Amanda time-out

2003-03-06 06:13:55
Subject: Amanda time-out
From: Jeroen Heijungs <Jeroen.Heijungs AT Het-Muziektheater DOT nl>
To: amanda-users AT amanda DOT org
Date: Thu, 06 Mar 2003 10:43:36 +0100
Dear listmembers,

I have a little problem here which I cannot solve so far:
We have some 13 servers (all FreeBSD in various versions from 3.5 to 4.7)
which I backup using Amanda 2.4.2, there are servers "before" a firewall,
and servers "behind" a firewall. All servers are backed up fine, it really
is trouble free, except for one of the servers behind the firewall. That
one (mailrelay server) was backed-up fine untill it became a production
server, suddenly it had every night a time-out error: 

"mailrelay. /var lev 0 FAILED [Request to mailrelay.xxxx.xx timed out.]"
for all filesystems.

At first I suspected the resolver or DNS, because this kind of problems
happened before by a wrong resolver, but that was not the case, also no
secondary IP adresses at the nic, after that I checked and increased the
timeout parameters: 

- etimeout is now at 2400 
- ctimeout is now at 120
amandad.debug started at 02:40:01, and ended at 02:41:25 at that server so
that should be enough, but the only effect so far is that the whole job is
taking an awfully long time to complete.

There are only two thing that seem to help:
- leave out the biggest file system (/dev/ad0s1f    24G   738M    21G
3%    /usr), then everything runs smoothly
- back up the same server two times (1-time by name, 2-time by IP-adress),
the first (name or IP-adress) to appear in the "disklist" is going well,
the second gets the error (??? that I do not understand, I stumbled upon
this when I tried the resolver and DNS solutions)

The error is always the same in amandad.debug:
Amanda 2.4 REP HANDLE 009-80FD0708 SEQ 1046914809
OPTIONS maxdumps=1;
/etc 0 SIZE 1260
/etc 1 SIZE 30
/home 0 SIZE 15710
/home 1 SIZE 40
/usr/local 0 SIZE 63730
/usr/local 1 SIZE 11610
/var 0 SIZE 9270
/var 1 SIZE 5020
----

amandad: waiting for ack: timeout, retrying
amandad: waiting for ack: timeout, retrying
amandad: waiting for ack: timeout, retrying
amandad: waiting for ack: timeout, retrying
amandad: waiting for ack: timeout, giving up!
amandad: pid 75481 finish time Thu Mar  6 02:41:25 2003



Perhaps someone has a bright idea??

tia
Jeroen Heijungs
Het Muziektheater
Amsterdam, The Netherlands


<Prev in Thread] Current Thread [Next in Thread>