IPv6 stack freeze caused by full ip6_dst_cache

This document (7018525) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 11
SUSE Linux Enterprise Server 11 Service Pack 1 (SLES 11 SP1)
SUSE Linux Enterprise Server 11 Service Pack 2 (SLES 11 SP2)
SUSE Linux Enterprise Server 11 Service Pack 3 (SLES 11 SP3)
SUSE Linux Enterprise Server 11 Service Pack 4 (SLES 11 SP4)
IPv4 and IPv6 setup

Situation

Systems running SLES 11 with active IPv4 and IPv6 configurations may encounter a situation resulting in loss of the IPv6 connection. The system however is accessible via IPv4.

Resolution

Workaround:

Please adjust the following value to prevent the IPv6 destination cache overflow.

echo 0 > /proc/sys/net/ipv4/tcp_tw_recycle

For reboot persistent settings please edit /etc/sysctl.conf and add:

net.ipv4.tcp_tw_recycle = 0

Resolution:

Please update to kernel 3.0.101-94.1 or later. The patches involved to address this issue are:

- tcp: fix inet6_csk_route_req() for link-local addresses (bsc#1010175).
- tcp: pass fl6 to inet6_csk_route_req() (bsc#1010175).
- tcp: plug dst leak in tcp_v6_conn_request() (bsc#1010175).
- tcp: use inet6_csk_route_req() in tcp_v6_send_synack() (bsc#1010175).

Cause

When a system responds to a TCPv6 request and time-wait sockets are recycled for new connections the table entries grow and lead to a table overflow, resulting in a freeze of the IPv6 stack.

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.