Operation Manual

SFX SERIES USER’S GUIDE
Rev 2.2 117
In addition to Device Type and Group filtering, 256 Priority levels are supported. The preferred master device should
be assigned the highest possible priority, generally up to 254. The highest priority 255 is a special case priority that
should only be used if you do not want the master device to ever give up its master status (including in non-preempt
mode – see below).
A priority of 0 is used by the current master device to give up its master status. If a master device fails or gives up, the
backup device with next highest priority will become master.
In addition to priorities, receivers can also be configured for Preempt Mode or non-Preempt Mode. In Preempt Mode, if
a device detects that the current master device has a lower priority than itself, it will become master. In non-Preempt
Mode, if a device detects that current master device has a lower priority than itself, it will not take over master
responsibilities. This mode can reduce service interruptions due to switch-overs. These modes are ignored if a device
is set to a priority of 255.
The length of time a backup receiver waits for a redundancy packet from the master before timing out and assuming
master status is:
Time (microseconds) = Advertisement Interval*1000000 + (256-Backup Priority)*100000;
So, if the backup has a priority of 253, the timeout will be 1.3 seconds with an Advertisement Interval of 1 second. If
the master receiver is highly utilized, or if there is a high probability of collision on the redundancy LAN, it is possible for
a redundancy advertisement to be delayed sufficiently for the backup to assume master status for a brief period of
time. In environments where this problem occurs, decreasing the priority of the backup unit can eliminate it.
For example, with an Advertisement Interval of 1 second, a Backup Priority of 245 would give a timeout of 2.1 seconds,
which would require that the backup miss two advertisements from the master before assuming master status. The
disadvantage is that if the master has really failed, the backup will not take over for 2.1 seconds (worst case).
N to M Redundancy
N to M Redundancy can be enabled in the Web GUI (see section later on in this Appendix). If it is disabled, then the
legacy 1:1 redundancy is assumed. N to M Redundancy allows one or more backup receivers to act as backups for
one or more master receivers, on the same LAN. In N to M Redundancy there are some important rules that have to
be followed:
1. Each master receiver must have its own group identity, and priority. (Master receivers should not share
Identities).
2. Backup receivers must be defined as N to M backup and have their own group identities (Note: all the
backup receivers can share group identities, unlike the master receivers). The backup receiver's priority
should be lower than all the master receivers in the network.
3. Preempt mode should be off, so that the Set Master and Set Backup functions will work properly.
4. The backup receiver does not have to be locked to a carrier, but it can be. If it is locked, all data will be
muted.
5. A single backup receiver can backup master receivers on different transponders. When a master receiver
loses master status (in the same way as described before), the backup receiver (with the highest priority) will
achieve master status and obtain the master’s configuration. This configuration includes: Carrier/LNB
settings, PID lists, routing tables, NCC, identity etc. The backup receiver will keep its own Ethernet port IP
addresses, however, to avoid network conflicts.
6. With preempt mode off, a command forcing a receiver back to master status has to be sent after it has
experienced a fault causing it to lose master status. This gives operator control of when the switchback to
master occurs.