Amdump fails to backup large DLEs: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
Paul.bijnens (talk | contribs) (not data connection but message connection is broken) |
||
Line 1: | Line 1: | ||
;Issue : [[amdump]] fails to backup large DLEs. It fails with "Connection reset by peer" error | ;Issue : [[amdump]] fails to backup large DLEs. It fails with "Connection reset by peer" error message | ||
message | |||
;Configuration : iptables (firewall packet filter) is configured on the Amanda | ;Configuration : iptables (firewall packet filter) is configured on the Amanda client | ||
;Cause : iptables was breaking the TCP connections | ;Cause : iptables was breaking the TCP connections that was set up between the Amanda client and server for transferring the error output of the backup command because it was idle. During backup of a large DLE, the message connection can remain idle for a long time. | ||
;Solution : Increase tcp keepalive probes | ;Solution : Increase tcp keepalive probes: | ||
echo 90 > /proc/sys/net/ipv4/tcp_keepalive_time | echo 90 > /proc/sys/net/ipv4/tcp_keepalive_time |
Revision as of 21:33, 5 December 2005
- Issue
- amdump fails to backup large DLEs. It fails with "Connection reset by peer" error message
- Configuration
- iptables (firewall packet filter) is configured on the Amanda client
- Cause
- iptables was breaking the TCP connections that was set up between the Amanda client and server for transferring the error output of the backup command because it was idle. During backup of a large DLE, the message connection can remain idle for a long time.
- Solution
- Increase tcp keepalive probes:
echo 90 > /proc/sys/net/ipv4/tcp_keepalive_time