Allied Telesis AR800 Series User Manual
Page 11

Patch 86231-10 For Rapier Switches and AR800 Series Modular Switching Routers
11
Patch 86231-10 for Software Release 2.3.1
C613-10328-00 REV J
When acting as a PPPoE Access Concentrator (AC), if a PPPoE client sent
discovery packets without the "host-unique" tag, the discovery packets sent
by the AC were corrupted. This issue has been resolved.
When an IP Multihomed interface was used as an OSPF interface,
neighbour relationships were only established if the IP interface for OSPF
was added first in the configuration. Now, OSPF establishes neighbour
relationships regardless of the IP Multihomed interface configuration order.
This patch implements simple DHCP range MIBs,
dhcpRangeExhaustedGateway and dhcpRangeExhaustedInterface, and a trap
that shows when a DHCP range was exhausted. The trap is triggered when
a DHCP request cannot be satisfied. The gateway address and the interface
address are returned as trap variables.
If the HELLOINTERVAL parameter in the SET PIM command was less than
4 seconds, the PIM neighbour would time out. The minimum
HELLOINTERVAL has been changed to 10 seconds for this command. Also,
the commands SET IP INTERFACE and SET PIM INTERFACE were not
recognising a modified IP address. The RESET PIM INTERFACE=VLAN
command was not working correctly. These issues have been resolved.
The Flash compaction message has been changed from:
Flash compacting...
DO NOT restart the router until compaction is completed
To:
Flash compacting...
DO NOT restart the router, or power down until compaction is completed
Address learning on the mirror port is now correctly re-enabled when it is
restored to its normal state.
If a gigabit fibre uplink was installed in a Rapier 24i or Rapier 48i, and the
receive fibre was slowly withdrawn while the uplink was receiving heavy
traffic, the switch chip sometimes locked up. This patch implements a
workaround by causing a warm restart if this situation arises. A message is
then written to the log to explain the cause of the restart.
IGMP failed to create an automatic IGMP membership with no joining port
when it received multicast data that no ports were interested in, when IP
TimeToLive was set to 1 second. Also, IGMP erroneously sent a query on an
IGMP enabled IP interface even when IGMP was disabled. These issues
have been resolved.
PCR: 02343
Module: PPP
Network affecting: No
PCR: 02359
Module: IPG
Network affecting: No
PCR: 02360
Module: DHCP
Network affecting: No
PCR: 02362
Module: PIM
Network affecting: No
PCR: 02364
Module: FFS
Network affecting: No
PCR: 02365
Module: SWI
Network affecting: No
PCR: 02366
Module: SWI
Network affecting: No
PCR: 02368
Module: IPG/IGMP
Network affecting: No