Brocade Multi-Service IronWare Switching Configuration Guide (Supporting R05.6.00) - Manual (Page 545)

Multi-Service IronWare Switching Configuration Guide 515
53-1003036-02
Ethernet Ring Protection
15
When an ERP topology starts up, each ERN (in Init state) transmits a R-APS (NR). After start-up, the
behavior varies by assigned role.
Figure 127 shows the initialization process for an ERN.
FIGURE 127 Message exchange and actions during ERN initialization version 2
When the ring is in the Pending state, an ERN flushes the filtering database (FDB) if it receives any
of the following state requests:
• Signal-fail (SF)
• No request (NR), RPL Blocked (RB)
NOTE
ITU-T G.8032 Version 1 does not use a Pending state, so from the Protection state ERNs enter the
Idle state.
ERP timers
ERP provides various timers to ensure stability in the ring while a recovery is in progress or to
prevent frequent triggering of the protection switching. All of the timers are operator configurable.
• Guard timer — All ERNs use a guard timer. The guard timer prevents the possibility of forming a
closed loop and prevents ERNs from applying outdated R-APS messages. The guard timer
activates when an ERN receives information about a local switching request, such as after a
switch fail (SF), manual switch (MS), or forced switch (FS). When this timer expires, the ERN
begins to apply actions from the R-APS it receives. This timer cannot be manually stopped.
• Wait to restore (WTR) timer — The RPL owner uses the WTR timer. The WTR timer applies to the
revertive mode to prevent frequent triggering of the protection switching due to port flapping or
intermittent signal failure defects. When this timer expires, the RPL owner sends a R-APS (NR,
RB) through the ring.
• Wait to Block (WTB) timers — This wait-to-block timer is activated on the RPL owner. The RPL
owner uses WTB timers before initiating an RPL block and then reverting to the idle state after
operator-initiated commands, such as for FS or MS conditions, are entered. Because multiple
FS commands are allowed to co-exist in a ring, the WTB timer ensures that the clearing of a
single FS command does not trigger the re-blocking of the RPL. The WTB timer is defined to be
5 seconds longer than the guard timer, which is enough time to allow a reporting ERN to
transmit two R-APS messages and allow the ring to identify the latent condition.
When clearing a MS command, the WTB timer prevents the formation of a closed loop due to
the RPL owner node applying an outdated remote MS request during the recovery process.
RPL owner Non-RPL node RPL node
Init state Init state Init state
1 Blocks the RPL
2 Sends a R-APS (NR)
3 Enters the Pending state.
1 Blocks the left interface
2 Sends a R-APS (NR)
3 Enters the Pending state
1 Blocks the left interface
2 Sends a R-APS (NR)
3 Enters the Pending state
4. Starts the WTR timer
5. (After the WTR expires) stops
sending NR
6. Sends R-APS (NR, RB, DNF)
7. Enters the Idle state After receiving the (NR, RB, DNF) from
the RPL owner:
1
Unblocks the non-failed blocking
port
2 Stops sending (NR)
3 Enters the Idle state
After receiving the (NR, RB, DNF) from
the RPL owner:
1
Blocks the RPL port
2 Unblocks the other ports
3 Enters the Idle state