S.CON1M/MA System
19
13 - Fault Propagation and SLE mechanisms
The main purpose of fault propagation mechanisms implemented in media converters
is to allow real-time notification of any link failure to upper hierarchy devices (e.g.
switches/routers with redundant links) consequently activating an alternate path to
communicate with a mission critical site. However, such mechanisms implemented in
other vendors’ products, may suffer from an inherent drawback. These are typical bi-
directional mechanisms: fault propagates both from fiber to TP port and from TP to the
fiber port. If such devices are modern and implement FEF as well, then it may become
a lethal combination: while propagating the fault as intended, when fault (original link
break) is repaired (i.e. original link is restored), such systems will hang and will be
restored only by manual intervention.
The powerful fault-propagation mechanism implemented in the S.CON1M/MA prevents
such a deadlock. The fault propagation is bi-directional, (F
>T, and/or T > F)
In order to optimize the network, you must first decide what is the critical site needing
redundancy, consequently establish the general direction of fault propagation required,
finally set up accordingly.
The mostly common use is (at the central node) F>>T: propagation: failure of the fiber
link will cut the TP link connected to the router/switch, which in turn will activate an
alternate path to the remote site. (In our case the remote device attached to the
S.CON1M/MA)
While the FP mechanism is “fool-proof”, the safe implementation of FP system-wise
depends now on remote device (the remote device connected to the opposite end of
the F/O link). If such device does not have a FP feature (or it is disabled) then no
problem – the network will behave according to the setting of the S.CON1M/MA
(F>>T enabled) without of risk of hanging. If the remote device is equipped with a FP
mechanism make sure that it is NOT bi-directional, otherwise you risk hanging as
explained above. In such case, it is recommended to configure the remote device to
one FP mode (F > T)
If the remote device is equipped with selective or just one-way FP then there should be
no risk in implementation. However, as such mechanisms are proprietary by nature,
we cannot guarantee end-to-end smooth operation.
If the remote device is another FibroLAN’s MA™ (like the H.CON/MA) and the Master
Unit (
S.CON1M/MA
,
FibroLAN Metrostar™, F.CON8/16/MA, F.CON12/MA/TL) is
managed, the most powerful SLE feature comes into the picture, further enhancing
your network’s resilience.
FibroLAN’s
S
ubscriber
L
ink
E
mulation virtually emulates the subscriber’s device or
network connected to the TP port of the remote device (converter) onto the TP link
of the Master Unit (connected usually to a switch/router). When activated
(
controllable
via the management system only)
it senses the loss of the TP link
connected to the remote MA device, and will cause the Master Unit to cut off its related
TP link (connected to a switch/router), alerting in real time the switch/router of the
subscriber failure.
All this is performed without disrupting the F/O link, thus
ensuring on-going control of the remote MA device via MA management.
