Specifications

ALAXALA Ring Configuration Guide (Edition 1)
Copyright © 2008, ALAXALA Networks Corporation. All rights reserved.
71
in fallback state due to the failure.
5.2 Parameter recommendations for chassis and box-type switch mixed networks
If you configure a ring that contains both chassis and box-type switches, you need to decide your parameter
values based on the performance of your box-type switches. This also applies to the case where you use a
chassis-type switch as the master node and box-type switches as transit nodes.
Table 5.2-1 shows the recommended parameter values. For an environment in this table, a normal ring and a
ring with LA are configured, chassis and box-type switches are mixed in those rings, and the shortest value is set
to the health-check interval parameter. The values in this table apply to a small-sized single ring. If you want to
build a large-scale ring network, you might need to tune up your network based on the table below.
Table 5.2-1 Parameter settings for mixed configuration of chassis- and box-type switches
Health-check holdtime [ms] Health-check
interval [ms]
Minimum Recommended
link debounce
time [ms]
UTP 500 750 1,500 --
Single
OPT 500 750 1,500 0
UTP 500 1,500 2,200 --
LA
OPT 500 1,200 2,000 0
UTP: for 10/100/1000BASE-T, OPT: for 1000BASE-X or 10GBASE-R
In Table 5.2-1 above, the health-check interval parameter has the minimum value. If you want to set a larger
value, you need to set a larger health-check holdtime value than the value in the table.
Also, if LA is not used, the recommended health-check holdtime value is set to more than three times the value
of the health-check interval value. If LA is used, you need to add a little extra time in order to maintain the ring
when the LA falls back.
(health-check holdtime) >= (health-check interval) x 3
When you set the minimum health-check holdtime value shown in Table 5.2-1 above, the time from the
occurrence of a ring failure to the failure detection becomes shorter. However, the master node might determine
that a ring failure has occurred in some cases, such as when HC frames are discarded on the way due to some
reason, in spite of the fact that the ring is in normal operation. When the recommended value is set, the master
node will not erroneously detect a ring failure if one of the HC frames is discarded because the longer
health-check holdtime value allows the master node to receive the subsequent HC frame. Instead, this longer
health-check holdtime value results in longer detection time of a ring failure.
5.3 Parameter settings when using the Ring Protocol with STP
If you use the Ring Protocol and STP together, there is no problem when you use the initial values for the ring
and STP to run both protocols. However, when you tune your parameters, you need to tune the parameters for
both the ring network and STP domain. Specifically, the following relationships must be satisfied:
(1) Ring parameter setting
(health-check holdtime) >= (health-check interval) x 3
(2) STP parameter setting
((forward-time) - 1) x 2 >= (max-age) >= (hello-time + 1) x 2
(3) Relationship between the ring parameter and STP parameter
(Ring's health-check holdtime) < (STP's hello-time x 3)
(4) Parameter settings to avoid a loop
Ring's forwarding-shift-time <= 10 sec
STP's spanning-tree single hello-time <= 2 sec
STP's spanning-tree vlan hello-time <= 2 sec