Bandwidth-based admission control mechanism
View Patent ↗A device generates a service protection factor (SPF i ) for links (N) on a link aggregation group (LAG), determines a traffic allocation bandwidth on the LAG for a service class based on the service protection factor (SPF i ) and a congestion guard factor (CGF i ), and restricts the traffic allocation bandwidth on the LAG. The service protection factor (SPF i ) may guarantee a service availability target in the presence of link failure, and the congestion guard factor (CGF i ) may protect against link overload that may arise from imperfectness in traffic load balancing across links in the LAG.
1. A computing-device implemented method, comprising:
generating, by the computing device, a service protection factor (SPF i ) for links (N) on a link aggregation group (LAG), where generating a service protection factor (SPF i ) includes:
calculating, by the computing device, a relationship between a data delivery ratio after link failure (LF_DDR i ) associated with a time period (T), and a mean time to repair failed links (T r ),
calculating, by the computing device, a bandwidth allocated for a service class on the LAG before link failure (BF_B i ),
calculating, by the computing device, a bandwidth allocated for the service class on the LAG after link failure (LF_B i ), and
determining, by the computing device, the service protection factor (SPF i ) based on the calculations;
determining, by the computing device, a traffic allocation bandwidth on the LAG for the service class based on the service protection factor (SPF i ); and
restricting, by the computing device, the traffic allocation bandwidth on the LAG.
2. The method of claim 1 , further comprising:
allocating bandwidth on the LAG based on the restricted traffic allocation bandwidth.
3. The method of claim 1 , where calculating a relationship includes:
calculating the relationship based on (T r ), (BF_B i ), (LF_B i ), (LF_DDR i ), and (T).
4. The method of claim 3 , where calculating a bandwidth allocated for the service class on the LAG before link failure (BF_B i ) includes:
calculating (BF_B i ) based on (SPF i ), a congestion guard factor (CGF i ), a raw link capacity before link failure (L), an allocated bandwidth for Ethernet Relay Service-Real Time (ERS-RT) traffic (B 0 ), and a weight assigned to the service class relative to other classes (w i ).
5. The method of claim 4 , where calculating a bandwidth allocated for the service class on the LAG after link failure (LF_B i ) includes:
calculating (LF_B i ) based on (L), a bookable bandwidth for ERS-RT traffic (U 0 ), (N), (w i ), and (CGF i ).
6. The method of claim 5 , where determining the service protection factor (SPF i ) based on the calculations includes:
determining the service protection factor (SPF i ) based on (T r ), (L), (U 0 ), (N), (B 0 ), (CGF i ), (I), and (LF_DDR i ).
7. The method of claim 1 , further comprising:
determining or receiving oversubscription guidelines.
8. The method of claim 1 , where determining a traffic allocation bandwidth includes:
determining the traffic allocation bandwidth for a class (i) based on an original bandwidth allocated for the service class (B i ), (SPF i ), a LAG bandwidth (L), a fraction of raw link capacity allocated for the service class (P i ), and an oversubscription factor for the service class (O i ).
9. The method of claim 1 , where restricting the traffic allocation bandwidth includes:
restricting the traffic allocation bandwidth for a class (i) based on an original bandwidth allocated for the service class (B i ), (SPF i ), and a congestion guard factor (CGF i ).
10. A device, comprising:
a plurality of links; and
a processor to:
calculate a relationship between a data delivery ratio after link failure (LF_DDR i ) associated with a time period (T), and a mean time to repair failed links (T r ),
calculate a bandwidth allocated for a service class on a link aggregation group (LAG), defined from two or more of the plurality of links, before link failure (BF_B i ),
calculate a bandwidth allocated for the service class on the LAG after link failure (LF_B i ),
determine, based on the calculations, a service protection factor (SPF i ) for links (N) on the LAG,
determine a traffic allocation bandwidth on the LAG for the service class based on the service protection factor (SPF i ), and
restrict the traffic allocation bandwidth on the LAG.
11. The device of claim 10 , where the processor is further to:
allocate bandwidth on the LAG based on the restricted traffic allocation bandwidth.
12. The device of claim 10 , where the device comprises at least one of a computer, a router, a switch, a network interface card (NIC), a hub, or a bridge.
13. The device of claim 10 , where the processor is further to:
calculate the relationship according to the following equation:
T r *( BF — B i −LF — B i )=(1 −LF — DDR i )* T*BF — B i .
14. The device of claim 13 , where the processor is further to:
calculate (BF_B i ) according to the following equation:
BF — B i =SPF i *CGF i *( L−B 0 )* w i ,
where (CGF i ) is a congestion guard factor, (L) is a raw link capacity before link failure, (B 0 ) is an allocated bandwidth for Ethernet Relay Service-Real Time (ERS-RT) traffic, and (w i ) is a weight assigned to the service class relative to other classes.
15. The device of claim 14 , where the processor is further to:
calculate (LF_B i ) according to the following equations:
LF_B
i
=
(
L
-
U
0
-
L
N
)
*
w
i
,
if
N
=
2
,
and
LF_B
i
=
(
L
--
U
0
-
L
N
)
*
w
i
*
CGF
i
,
if
N
>
2
,
where (U 0 ) is a bookable bandwidth for ERS-RT traffic.
16. The device of claim 15 , where the processor is further to:
determine the service protection factor (SPF i ) according to the following equations:
SPF
i
=
T
r
*
(
L
-
U
0
-
L
N
)
(
L
-
B
0
)
*
CGF
i
*
(
T
r
-
T
+
LF_DDR
i
*
T
)
,
if
N
=
2
,
and
SPF
i
=
T
r
*
(
L
-
U
0
-
L
N
)
(
L
-
B
0
)
*
(
T
r
-
T
+
LF_DDR
i
*
T
)
,
if
N
>
2.
17. The device of claim 10 , where the processor is further to:
determine or receive oversubscription guidelines.
18. The device of claim 10 , where the processor is further to:
determine the traffic allocation bandwidth for a class (i) according to the following equations:
U i ≦B i *SPF i , where 0< SPF i ≦1 and B i =L*P i *O i , and
where (B i ) is an original bandwidth allocated for the service class, (L) is a LAG bandwidth, (P i ) is a fraction of raw link capacity allocated for the service class, and (O i ) is an oversubscription factor for the service class.
19. The device of claim 10 , where the processor is further to:
restrict the traffic allocation bandwidth for a class (i) according to the following equations:
U i ≦B i *SPF i *CGF i , where 0 <CGF i ≦1, and
where (B i ) is an original bandwidth allocated for the service class, and (CGF i ) is a congestion guard factor.
20. A system, comprising:
means for generating a service protection factor (SPF i ) for links (N) on a link aggregation group (LAG), where the means for generating includes:
means for calculating a relationship between a data delivery ratio after link failure (LF_DDR i ) associated with a time period (T), and a mean time to repair failed links (T r ),
means for calculating a bandwidth allocated for a service class on the LAG before link failure (BF_B i ),
means for calculating a bandwidth allocated for the service class on the LAG after link failure (LF_B i ), and
means for determining the service protection factor (SPF i ) based on the calculations;
means for determining a traffic allocation bandwidth on the LAG for the service class based on the service protection factor (SPF i );
means for restricting the traffic allocation bandwidth on the LAG; and
means for allocating bandwidth on the LAG based on the restricted traffic allocation bandwidth.