
456
Fabric OS Administrator’s Guide
53-1002446-01
Buffer credit management
23
Upon arrival at a receiver, a frame goes through several steps. It is received, deserialized, decoded,
and is stored in a receive buffer where it is processed by the receiving port. If another frame arrives
while the receiver is processing the first frame, a second receive buffer is needed to hold this new
frame. Unless the receiver is capable of processing frames as fast as the transmitter is capable of
sending them, it is possible for all of the receive buffers to fill up with received frames. At this point,
if the transmitter should send another frame, the receiver will not have a receive buffer available
and the frame is lost. Buffer-to-Buffer flow control provides consistent and reliable frame delivery of
information from sender to receiver.
Optimal buffer credit allocation
The optimal number of buffer credits is determined by the distance (frame delivery time), the
processing time at the receiving port, link signaling rate, and size of the frames being transmitted.
As the link speed increases, the frame transmission time is reduced and the number of buffer
credits must be increased to obtain full link utilization, even in a short-distance environment.
For each frame that is transferred, the hardware at the other end must acknowledge that the frame
has been received before a successful transmission occurs. This requires enough capacity in the
hardware to allow continuous transmission of frames on the link, while waiting for the
acknowledgement to be sent by the receiver at the other end.
As the distance between switches and the link speed increases, additional buffer credits are
required for the ports used for long-distance connections. Distance levels define how buffer credits
are allocated and managed for extended ISLs. Buffer credits are managed from a common pool
available to a group of ports on a switch. The buffer credit can be changed for specific applications
or operating environments, but it must be in agreement among all switches to allow formation of
the fabric.
To maintain 100 percent utilization of a 1 Gbps link for 100 km, the sending hardware must have
enough resources (BB credits) to keep 106,250 bytes on the link and the receiving hardware must
have enough resources to allow the sender to transmit continuously. To theoretically achieve 100
percent utilization of a 2 Gbps link for 100 km, the required number of BB credits ranges from 98
to 2310 depending on the average frame size. When the link speed is increased to 4 Gbps, the
required number of BB credits ranges from 196 to 4620. It is not possible for the switch to
determine what the frame size is going to be.
Considerations for calculating buffer credits
Following are the considerations for calculating how many ports can be configured for long
distance on all Fabric OS v7.x-capable switch modules:
•
Each port is part of a port group that includes a pool of buffer credits that can be utilized. This
is not the same as the port groups used for ISL Trunking.
•
Each user port reserves eight buffer credits when online or offline.
•
Any remaining buffers can be reserved by any port in the port group.
•
When QoS is enabled and the port is online, an additional 20 buffers are allocated to that port.
•
The FR4-18i blade has a limitation of 255 buffers maximum that can be allocated to a port,
which corresponds to a distance of ~500 km at 1 Gbps.
Содержание Fabric OS v7.0.1
Страница 1: ...53 1002446 01 15 December 2011 Fabric OS Administrator s Guide Supporting Fabric OS v7 0 1 ...
Страница 22: ...xxii Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 26: ...xxvi Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 30: ...xxx Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 38: ...xl Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 40: ...2 Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 100: ...62 Fabric OS Administrator s Guide 53 1002446 01 Configuring FLOGI time handling of duplicate PWWN 3 ...
Страница 152: ...114 Fabric OS Administrator s Guide 53 1002446 01 The authentication model using RADIUS and LDAP 5 ...
Страница 214: ...176 Fabric OS Administrator s Guide 53 1002446 01 Management interface security 7 ...
Страница 228: ...190 Fabric OS Administrator s Guide 53 1002446 01 Brocade configuration form 8 ...
Страница 248: ...210 Fabric OS Administrator s Guide 53 1002446 01 Validating a firmware download 9 ...
Страница 334: ...296 Fabric OS Administrator s Guide 53 1002446 01 Setting up TI over FCR sample procedure 12 ...
Страница 360: ...322 Fabric OS Administrator s Guide 53 1002446 01 Encryption and compression example 14 ...
Страница 374: ...336 Fabric OS Administrator s Guide 53 1002446 01 Access Gateway N_Port failover with FA PWWN 16 ...
Страница 404: ...366 Fabric OS Administrator s Guide 53 1002446 01 ...
Страница 430: ...392 Fabric OS Administrator s Guide 53 1002446 01 Ports on Demand 18 ...
Страница 470: ...432 Fabric OS Administrator s Guide 53 1002446 01 Disabling QoS zone based traffic prioritization 21 ...
Страница 502: ...464 Fabric OS Administrator s Guide 53 1002446 01 Buffer credit recovery 23 ...
Страница 544: ...506 Fabric OS Administrator s Guide 53 1002446 01 Displaying the range of output ports connected to xlate domains 24 ...
Страница 554: ...516 Fabric OS Administrator s Guide 53 1002446 01 Fabric configurations for interconnectivity A ...
Страница 572: ...534 Fabric OS Administrator s Guide 53 1002446 01 Hexadecimal overview D ...
Страница 584: ...546 Fabric OS Administrator s Guide 53 1002446 01 ...