JunosE 11.0.2 Release Notes
20
Known Behavior
DHCP external server may not be able to bind all DHCP clients when all of the
following conditions exist:
DHCP external server and either DHCP relay or relay proxy are configured
in separate virtual routers on an E320 router.
The client-facing and server-facing interfaces for DHCP external server and
either DHCP relay or relay proxy are configured on the same ES2 4G LM.
DHCP external server is configured to create dynamic subscriber
interfaces.
When these three conditions exist simultaneously, the ES2 4G LM may not be
able to successfully process all DHCP packets. Although all clients may get
bounded in DHCP relay or relay proxy, some clients may not get bounded in
DHCP external server. (In a production environment it is highly unlikely for
conditions 1 and 2 to exist because stand-alone DHCP external server is
normally configured for a DHCP relay in a different chassis.)
Work-around:
You can eliminate this issue by modifying any one of these
conditions. For example, this issue does not exist with any of the following
configuration modifications:
Configure DHCP external server and either DHCP relay or relay proxy in
the same virtual router.
Configure the client-facing and server-facing interfaces for DHCP external
server and either DHCP relay or relay proxy on the same ES2 10G LM
instead of the same ES2 4G LM.
Configure the client-facing and server-facing interfaces for DHCP external
server and either DHCP relay or relay proxy on separate ES2 4G LMs.
DHCP NAK packets are sent from a different VLAN than the one on which the
renew request is received on a router that is configured with dynamic VLANs,
DHCP local server, and automatically created dynamic subscriber interfaces.
This behavior occurs only after a link flap has taken place. [Defect ID 87062]
Ethernet
The hashing algorithm that selects the LAG member link is associated with the
IP address of the subscriber client to support QoS. Consequently, a particular
flow is always hashed to the same link. When a member link is removed from
a LAG bundle, traffic rate is disrupted and traffic flow is reduced. When the link
goes down and then comes back up, the traffic flow is automatically
redistributed.
When counting bits per second on a Fast Ethernet or Gigabit Ethernet interface,
an E Series router includes 12 bytes for interpacket gap, 7 bytes for preamble,
and 1 byte for start frame delimiter, for a total of 20 bytes (160 bits) per packet
more than some non–E Series routers. This value therefore shows the total
bandwidth utilization on the interface, including both data and overhead.