Hello we have a amanda set up that have been running for a long time
working like a clock, but since I upgraded the Linux kernel I've started
to see random disks failing with "[mesg read: Connection reset by
peer]". Which disk that fail varies from night to night (and some nights
everything works without an hitch).
I can't find anything special in the server logs, (except for some minor
network driver problems the first night, that I solved). On the failing
clients there is entries along the lines (somewhat sanitized):
---
/usr/local/libexec/sendbackup: version 2.4.3b3
sendbackup: got input request: DUMP <diskname> 0 1970:1:1:0:0:0 OPTIONS
|;bsd-auth;compress-fast;index;
parsed request as: program `DUMP'
disk `<diskname>'
device `<diskname>'
lev 0
since 1970:1:1:0:0:0
opt `|;bsd-auth;compress-fast;index;'
sendbackup: try_socksize: send buffer size is 65536
sendbackup: stream_server: waiting for connection: 0.0.0.0.1137
sendbackup: stream_server: waiting for connection: 0.0.0.0.1138
sendbackup: stream_server: waiting for connection: 0.0.0.0.1139
waiting for connect on 1137, then 1138, then 1139
sendbackup: stream_accept: connection from <server ip>.1124
sendbackup: stream_accept: timeout after 30 seconds
sendbackup: timeout on mesg port 1138
sendbackup: stream_accept: connection from <server ip>.1171
sendbackup: pid 75056 finish time Mon Sep 11 23:45:52 2006
---
I know that 2.4.3b3 is old but I'd like to avoid changing version until
some real world things have fallen into place.
The failing clients have been of all kinds (AIX 4.3, Linux (kernel 2.2,
2.6) and even the single hpux we run) and mostly situated on the same
lan (no firewall).
Hoping someone can prod me in the right direction.
/jp
--
jens persson # Cthulhu for president
<jens AT persson DOT cx> # Why vote for a lesser evil?
Mäster Olofsväg 24 # -- dlc (on kuro5hin.org)
S-224 66 LUND;SWEDEN #
|
