Released
PMC-Sierra, Inc.
PM9311/2/3/5 ETT1™ CHIP SET
Data Sheet
PMC-2000164
ISSUE 3
ENHANCED TT1™ SWITCH FABRIC
1.9.6.3 Scheduler Board
In the non-redundant Scheduler configuration, a failed Scheduler board will prevent all traffic from flowing
in the entire system. The following steps are required to cope with this failure.
1. Detect failed Scheduler board. For example, CPU is indicated of loss of connectivity from this
board to any one of the existing port boards.
2. Disable TDM and non-TDM traffic to/from ports that still have connectivity.
3. Disable the AIB links to/from the Port board.
4. Replace failed Scheduler board.
5. Bring up new board (see “Scheduler Board being added:” on page 89)
6. Configure this Scheduler to operate in non-redundant Scheduler mode.
7. Enable TDM traffic in Scheduler for connected ports. TDM traffic can now flow.
8. Refresh state information from all ports to this Scheduler. (See Section 1.9.4 “Scheduler Refresh
Procedure”)
9. Best-effort traffic can now flow.
The steps required for a redundant Scheduler configuration are more complicated. The redundant
Scheduler can still schedule traffic while the failed Scheduler board is replaced. When the new Scheduler
is inserted into the system, it must be resynchronized with the other Scheduler and the Enhanced Port
Processors. This will be accomplished by the following steps.
1. Detect failed Scheduler board. For example, CPU is indicated of loss of connectivity from this
board to any one of the existing port boards. This failure will cause this Scheduler to be disabled.
2. Disable the AIB links to/from the Port board.
3. Replace failed Scheduler board.
4. Bring up new board (see “Scheduler Board being added:” on page 89)
5. Configure this Scheduler to operate in redundant Scheduler mode.
6. If the Scheduler is generating the TDM sync, then TDM traffic must be stopped by writing a value
of 0 to the TDM Sync Period and then writing the actual value to both Schedulers.
7. Enable TDM traffic in new Scheduler for appropriate ports.
NOTE: The redundant Scheduler is still scheduling TDM and non-TDM traffic.
8. Refresh state information from all ports to this Scheduler. (See Section 1.9.4 “Scheduler Refresh
Procedure”) TDM traffic still flows.
9. Best-effort traffic can now flow. Schedulers are now in fault tolerant mode and synchronized.
PROPRIETARY AND CONFIDENTIAL TO PMC-SIERRA, INC., AND FOR ITS CUSTOMERS’ INTERNAL USE
101
PMC [ PMC-SIERRA, INC ]
