Share or add to bookmark

AddThis Social Bookmark Button

Latest comments

Authorization



  
Протоколы
Обзор протоколов передачи данных

1



SRDF with SiRT

SRDF with SiRT
SiRT (Single RoundTrip) for Fibre Channel SRDF directors (RFs) was introduced in Enginuity 5772 for SRDF/S mode only. It is dynamically enabled for SRDF/S links > 12 Km for Glossary Link block sizes up to 32K in Enginuity 5773 code. SiRT is compatible with fast write/write acceleration switches and extenders, as it will measure link latency and disable automatically if connected to these devices. As a best practice, it is recommended that either the EMC SiRT feature or the third-party fast write feature should be used. Both should not be enabled simultaneously.
The Fibre Channel SiRT feature for the Fibre Channel director can be set to Off or Automatic. When set to Automatic, this feature will only accelerate write I/Os using criteria based on latency and I/O size.

Note: EMC recommends contacting your EMC Customer Service Representative to verify that the setting is enabled if required in your environment.

Figure 5 shows the normal write process without the SiRT feature.
 Figure 5 Normal write command process
Figure 5 Normal write command process

The intended purpose of this feature is to maintain SRDF/S synchronicity while improving performance by localizing the transfer-ready response to the local RF port, thereby reducing an unnecessary acknowledgement response (trip) over the dark fiber distance (step 2 in Figure 5). Immediate benefits are apparent upon activation in transparent SRDF synchronous distance extension environments.
If applicable, multiple SRDF synchronous links can maximize their I/O performance over the Glossary Link network (transparent WDM environment).
In the example shown in Figure 6, RF1 (R1 F_Port) and RF2 (R2 F_Port) are managing the SiRT flow control.
 Figure 6 SRDF SiRT
Figure 6 SRDF SiRT
Legend:

Red

RF-ports with SiRT activated.

Blue

A step-by-step of a single write command with SiRT enabled.

 

Fast write/ write acceleration
EMC Connectrix and other third-party products offer single roundtrip for Fibre Channel capabilities (fast write/write acceleration) that can also increase SRDF throughput for direct-attach or Fibre Channel switched fabric configurations over extended distances. It is transparent to SRDF FC links and is used for all SRDF modes to decrease response time (SRDF/S) or improve performance over long distance links (mostly for adaptive copy and SRDF/AR, but also for some SRDF/A configurations).
Figure 7 shows a write command with fast write features.
 Figure 7 Write command with SiRT
Figure 7 Write command with SiRT

For Connectrix or third-party products, refer to the EMC Support Matrix available at http://elabnavigator.EMC.com to verify which of these products are supported for SRDF configurations.
! IMPORTANT
Not all products offering this feature are supported with SRDF due to unique write commands utilized by SRDF. SiRT with distance vendor write acceleration
With this in mind SiRT usage, in combination with the distance extension device-offered write acceleration mode, must be addressed.
Essentially for environments where the distance extension device is already servicing write commands on an E_Port level, it is recommended to disable SiRT. Refer to Figure 8.
 Figure 8 All F_Ports will benefit
Figure 8 All F_Ports will benefit
Legend:

Red

RF ports benefiting from distance extension device, write acceleration.

Blue

Scope.

 

In Figure 8, by enabling the write acceleration feature on the distance extension device, potentially all F_Ports (RF ports, FA ports, tape, etc.) issuing writes traversing across the E_Port attached to the distance extension client port can also take advantage of the throughput benefits from the activated write acceleration feature.

Link initialization
For link initialization of a Fibre Channel port, Fibre Channel specifications state that the maximum tolerable response time for a response is 100 milliseconds roundtrip time. This timeframe coincides with the limited timeframe of the Receiver-Transmitter Timeout Value (R_T_TOV), which is how long an FC port listens for a link response to a link service before an error is detected.

 
Distance extension consideration

Distance extension consideration
This section provides information to consider when using distance extension.
◆ “Link speed
◆ “Data buffering and flow control
◆ “TCP/IP window”
◆ “Active and passive devices”
◆ “FC SONET/GbE/IP”

Link speed
Link speed is an important aspect of distance extension configurations. Within the SAN networks link speeds equate to the amount of maximum Glossary Link bandwidth reachable on an E_Port and/or an F_Port. There are a variety of link speeds that are supported in a SAN Glossary Link network. Table 1 compares and contrasts the STS, optical carrier, and Fibre Channel link speed rates.
Table 1 STS-1s and optical carrier rates

 

STS

Optical carrier

Optical carrier rate

Fibre Channel link speeds

STS-1

OC-1

51.84 Mb/s

 

STS-3

OC-3

155.52 Mb/s

 

STS-12

OC-12

622.08 Mb/s

 

STS-24

OC-24

1244.16 Mb/s

1.0625 Gb/s or 100 MB/s

STS-48

OC-48

2488.32 Mb/s

2.125 Gb/s or 200 MB/s

STS-96

OC-96

4976.64 Mb/s

4.250 Gb/s or 400 MB/s

STS-192

OC-192

9953.28 Mb/s

10.51875 Gb/s or 12.75 Gb/s

STS-768

OC-768

39813.120 Mb/s

 

STS-3072

OC-3072

159252.240 Mb/s

 

 

Data buffering and flow control
The following information is discussed in this section:
◆ “Fibre Channel,” next
◆ “Maximum supported distance per Fibre Channel BB_Credit guidelines”
◆ “Buffer-to-buffer credit information”

Fibre Channel
Fibre Channel uses the BB_Credit (buffer-to-buffer credit) mechanism for hardware-based flow control. This means that a port has the ability to pace the frame flow into its processing buffers. This mechanism eliminates the need of switching hardware to discard frames due to high congestion. EMC testing has shown this mechanism to be extremely effective in its speed and robustness.
BB_Credit management occurs between any two Fibre Channel ports that are connected. For example:
◆ One N_Port and one F_Port
◆ Two E_Ports
◆ Two N_Ports in a point-to-point topology
◆ In Arbitrated Loop different modes
The standard provides a frame-acknowledgement mechanism in which an R_RDY (Receiver Ready) primitive is sent from the receiving port to the transmitting port for every available buffer on the receiving side. The transmitting port maintains a count of free receiver buffers, and will continue to send frames if the count is greater than zero.
The algorithm is as follows:
1. The transmitter's count initializes to the BB_Credit value established when the ports exchange parameters at login.
In an Arbitrated Loop environment the credits are established by the receiving port sending in advance R_RDY primitives after the login to establish the credit.
2. The transmitting port decrements the count per transmitted frame.
3. The transmitting port will stop sending frames when the credit reaches zero.
4. When a link reset occurs, the credit values are reestablished to values negotiated upon login.
5. The transmitting port increments the count per R_RDY it receives from the receiving port.
Figure 1 provides a view of the BB_Credit mechanism.
 Figure 1 BB_Credit mechanism
Figure 1 BB_Credit mechanism

As viewed from Port A’s perspective, when a link is established with Port B, BB_Credit information is exchanged. In this case, Port B provided a BB_Credit count of 5 to Port A. For Port A, this means it can transmit up to five Fibre Channel frames without receiving an R_RDY.

Maximum supported distance per Fibre Channel BB_Credit guidelines
In order to achieve maximum utilization of the Fibre Channel link it is highly advisable that both ports, connected on either side of the long haul setup provided by the DWDM, be capable of high BB_Credit counts. Use the following formula to calculate the approximate BB_Credit(s) required for the specific long haul application.
To calculate for BB_Credits, use the following formula for calculating the required BB_Credit count:

Speed

Formula

1 Gb/s

BB_Credit = ROUNDUP [2 * one-way distance in km/4] * 1

2 Gb/s

BB_Credit = ROUNDUP [2 * one-way distance in km/4] * 2

4 Gb/s

BB_Credit = ROUNDUP [2 * one-way distance in km/4] * 4

8 Gb/s

BB_Credit = ROUNDUP [2 * one-way distance in km/4] * 8

10 Gb/s

BB_Credit = ROUNDUP [2 * one-way distance in km/4] * 12


The factor of 2 in the formulas accounts for the time it takes the light to travel the entire roundtrip distance: frame from transmitter to receiver and R_RDY back to transmitter.
Maximum allowable distance is based on optical power measurements of the site. These measurements should be approved by DWDM and fiber services provider(s). The distance between an ISL ports on a Fibre Channel switch to a DWDM port should be included as part of the total distance (d1+d2+d3).

DWDM
The following BB_Credit charts will aid in providing estimates in regards to the amount of credits that should be present on the link when factoring Fibre Channel link speeds and link distances between the E_Ports.
Assuming the following is true:
◆ Light propagation in glass is 5 microsec/km, or 59 sec/m.
◆ Frame size is 2148 bytes/frame.
◆ Fibre Channel bit rate depends on the Fibre Channel speed.
Maximum distances assume 100% utilization of the ISL. If the ISL is not fully utilized, greater distances can be achieved since more BB_Credits become available. For example, for a 2 Gb/s switch port with 120 BB_Credits and with an ISL that is only 50% utilized, the maximum distance is 240 km.
Since Brocade’s credit information is provided by ASIC types, review Table 2 to correlate between switch ASIC and model numbers.

Table 2 Switch ASIC and model numbers

Vendor

ASIC/Family

EMC name

Vendor name

Brocade

Loom

Connectrix DS-16B

SilkWorm 2800

Bloom

Connectrix DS-16B2

SilkWorm 3800

Bloom

Connectrix DS-32B2

SilkWorm 3900

Bloom

Connectrix ED-12000B

SilkWorm 12000

Bloom2

Connectrix ED-24000B

SilkWorm 24000

Bloom2

Connectrix DS-16B3

Silkworm 3850

Bloom2

Connectrix DS-8B3

SilkWorm 3250

Condor

Connectrix DS-4100B

SilkWorm 4100

Condor

Connectrix ED-48000B

SilkWorm 48000

Condor

Connectrix DS-4900B

SilkWorm 4900

Goldeneye

Connectrix DS-220B

SilkWorm 200

Cisco

Vegas (MDS generation 1)

 

9509

 

 

9506

 

 

9216

 

 

9216A

 

 

9216i

 

 

9120

 

 

9140

Isolla (MDS generation 2)

 

9513

Brocade M-Series

Stitch

ED-1032

ED-5000

Viper / Fuji-Shasta

DS-16M

ES-3016

 

DS-16M2

ES-3216

 

DS-32M

ES-3032

 

DS-32M2

ES-3232

 

ED-64M

ED-6064

 

ED-140M

ED-6140

Posideon/Teton

N/A

ES-4300

 

DS-24M2

ES-4500

Sanera

ED-10000M

Intrepid 10000

Pegasus/Teton

DS-4400M

ES-4400

 

DS-4700M

ES-4700

 
Distance extension technologies overview
Distance extension technologies overview
To comprehend the distance extension solutions for Storage Area Networks it is important to understand and recall the challenges when implementing SAN connectivity over remote distances. The following information is provided in this overview section:
◆ “Early implementations of SAN environments ,” next
◆ “DWDM
◆ “CWDM
◆ “SONET
◆ “GbE
◆ “TCP/IP

Early implementations of SAN environments
To increase a single port between two Fibre Channel switches separated by a large geographical distance, every two strands (transmit, receive) of optical fiber cable were required to be physically added by the distance provider. The customer would generally incur expensive construction, service, and maintenance costs when adding a bulk of fiber cables intended to satisfy current E_Port connectivity requirements while allowing future growth potential and redundancy against accidental fiber breaks. Existing fibers that were used for Ethernet implementations could not be shared and required separate dedicated channels per protocol. The challenges involved with this process would stem anywhere from mandatory to extraneous costs associated with fiber cable maintenance. In addition to costs, there were physical hardware limitations to achieving connectivity between (at least) two geographically separated sites. Fibre Channel optics installed on the Fibre Channel switch were at the mercy of the limited optical output transmission power. Even with repeater technology, distortion of the optical wavelength transmitted by the optics can occur over several hops.
The Fibre Channel switches provided limitations as well. Link initialization and flow control were solely controlled by the Fibre Channel switches. The Fibre Channel standard would actually dictate the thresholds in regards to supporting large distances through optical connectivity and the obtainable Glossary Link bandwidth between two Fibre Channel ports. To finalize the list of challenges that SAN environments had to overcome, each Fibre Channel switch provider had its own non-standard and standard ways of implementing their native environments. This may deviate from the mass interpretation of the Fibre Channel standards.

DWDM
Dense Wavelength Division Multiplexing (DWDM) is a process in which different channels of data are carried at different wavelengths over one pair of fiber-optic links. This is in contrast with a conventional fiber-optic system in which just one channel is carried over a single wavelength traveling through a single fiber.
Using DWDM, several separate wavelengths (or channels) of data can be multiplexed into a multicolored light stream transmitted on a single optical fiber (dark fiber). This technique to transmit several independent data streams over a single fiber link is an approach to opening up the conventional optical fiber bandwidth by breaking it up into many channels, each at a different optical wavelength (a different color of light). Each wavelength can carry a signal at any bit rate less than an upper limit defined by the electronics, typically up to several gigabits per second.
Different data formats being transmitted at different data rates can be transmitted together. Specifically, IP data, ESCON SRDF, Fibre Channel SRDF, SONET data, and ATM data can all be traveling at the same time within the optical fiber.
DWDM systems are independent of protocol or format, and no performance impacts are introduced by the system itself.

Figure 1 illustrates the DWDM technology concept:
 Figure 1 DWDM example
Figure 1 DWDM example

For EMC customers it means that multiple SRDF® channels and Fibre Channel ISL (Inter Switch Links) can be transferred over one pair of fiber links along with traditional Glossary Link network traffic. This is especially important where fiber links are at a premium. For example, a customer may be leasing fiber, so the more traffic they can run over a single link, the more cost effective the solution.
With today's technology, the capacity of a single pair of fiber strands is virtually unlimited. The limitation comes from the DWDM itself. Optical-to-electrical transfers for switching and channel protection are required and limit the input traffic per channel.
Available DWDM topologies include point-to-point and ring configurations with protected and unprotected schemas. DWDM technology can also be used to tie two or more metro area data centers together as one virtual data center.
DWDM systems can multiplex and de-multiplex a large amount of channel quantities. Each channel is allocated its own specific wavelength (lambda) band assignment. Each wavelength band is generally separated by 10 nm spacing(s). As optical technologies improve, separations between each channel may be further reduced enabling more channels to be packed (tighter) onto a single duplex dark fiber.
DWDM has a higher cost associated due to greater channel consolidation, flexibility, utilization of higher quality hardware precision-cooling components (to prevent low frequency signal drift) and the capabilities of regenerating, re-amplifying and reshaping (3R) wavelengths assigned to channels to ensure optical connectivity over vast distances.
Varying circuits pack capabilities are also offered in a DWDM environment. DWDM circuit packs / blades can provide the following protocol conversions:
◆ Fibre Channel to SONET
◆ Fibre Channel to Gigabit Ethernet
◆ Fibre Channel to IP
In addition, some circuit packs can enable features such as write acceleration and buffer-to-buffer credit spoofing. To verify the latest supported distance systems and features, refer to the EMC Support Matrix.
Figure 2 shows a general concept of Fibre Channel link extension using DWDM.
 
d1 = DWDM signal over dark fiber medium.
d2 and d3 = Local ISL connections between switches and DWDM input.
Can be SM or MM depending on DWDM and switch interfaces or local distance requirements.
d4 and d5 = Local storage or server connections into the fabric.
 
 Figure 2 Fibre Channel link extension
Figure 2 Fibre Channel link extension

Note: All components are randomly selected and do not reflect a specific setup or configuration.
Note: Distance limitation may also be affected by application response time-out values and should consider signal propagation delay over site distance.
The following list provides general envelope guidelines for using DWDM systems:
◆ May be used for ESCON RDF distance extension, with direct connection between Symmetrix ESCON director ports and DWDM input ports.
◆ May be used for ISL extension of Fibre Channel switched fabrics. (E-Lab Navigator describes switch compatibility.)
◆ Fabric topology guidelines are provided per Fibre Channel switch topology documentation.
◆ Direct connections between host HBA or Symmetrix Fibre Channel director to a DWDM port are not supported. E-Lab Navigator contains specific DWDM distance and topology guidelines.
◆ As a general approach, two distances need to be measured. The shorter of the two is the maximum distance to be supported in the site.

CWDM
Coarse Wave Division Multiplexing (CWDM), like DWDM, uses similar processes of multiplexing and de-multiplexing different channels by assigning different wavelengths to each channel. CWDM is intended to consolidate environments containing a low number of channels at a reduced cost.
CWDM contains 20 nm separations between each assigned channel wavelength. CWDM technology generally uses cost-effective hardware components that require a reduced amount of precision-cooling components usually dominant in DWDM solutions due to the wider separations. With CWDM technology the number of channel wavelengths to be packed onto a single fiber is greatly reduced.
CWDM implementations, like DWDM, utilize an optical-to-electrical-to-optical technology where all the channels are multiplexed into a single CWDM device performing the optical-to-electrical-to-optical conversion.
A CWDM connectivity solution can use optics generating a higher wavelength with increased output optical power. Each channel is designated its own specific wavelength by the specific hot-pluggable CWDM GBIC/SFP optic installed on the Fibre Channel Switches.
With clean fibers, minimal patch panel connections, and ample optical power, CWDM optics alone can provide connectivity distances of up to 100 km per channel. To complete this solution a passive MUX/DEMUX is required to consolidate multiple channel-wavelengths into a single duplex 9-micron dark fiber.

Differences between DWDM and CWDM
The following are differences between DWDM and CWDM:
◆ Number of channels that are supported per solution.
DWDM systems can support channels ranging from 16 channels or above while CWDM supports 16 channels or below.
◆ CWDM GBIC/SFP optics can be used to increase the wavelength output of a channel (i.e., FC-switch optics).
The CWDM GBIC/SFP optics is usually installed in the Fibre Channel switch or client device. The wavelength and optical power enhanced links are then multiplexed and de-multiplexed to and from a single-mode 9-micron dark fiber.
◆ Costs.
Hardware components included with DWDM units are higher in cost due to precision-cooling techniques required to prevent signal drift. DWDM offers greater channel flexibility and capacity.
◆ Configurations can be complex with CWDM.
CWDM requires specific optics for each specific wavelength. Growth for a CWDM environment is limited and difficult to manage when supporting environments growing to larger channel support. More cabling would be required, thereby increasing complexity.
◆ DWDM devices offer circuit packs with numerous features (i.e., protocol conversions, buffer-to-buffer credit spoofing, write acceleration).

SONET
Synchronous Optical NETwork, (SONET), is a standard for optical telecommunications transport, developed by the Exchange Carriers Standards Association for ANSI. SONET defines a technology for carrying different capacity signals through a synchronous optical network. The standard defines a byte-interleaved multiplexed transport occupying the physical layer of the OSI model.
Synchronization is provided by one principal network element with a very stable clock (Stratum 3), which is sourced on its outgoing OC-N signal. This clock is then used by other network elements for their clocks (loop timing).
SONET is useful in a SAN for consolidating multiple low-frequency channels (Client ESCON and 1, 2 Gb Fibre Channel) into a single higher-speed connection. This can reduce DWDM wavelength requirements in an existing SAN infrastructure. It can also allow a distance solution to be provided from any SONET service carrier, saving the expense of running private optical cable over long distances.
The basic SONET building Glossary Link block is an STS-1 (Synchronous Transport Signal), composed of the transport overhead plus a Synchronous Payload Envelope (SPE), totaling 810 bytes. The 27-byte transport overhead is used for operations, administration, maintenance, and provisioning. The remaining bytes make up the SPE, of which an additional nine bytes are path overhead. It is arranged as depicted in Figure 3. Columns 1, 2, and 3 are the transport overhead.

 Figure 3 STS-1 organization
Figure 3 STS-1 organization

An STS-1 operates at 51.84 Mb/s, so multiple STS-1s are required to provide the necessary bandwidth for ESCON, Fibre Channel, and Ethernet, as shown in Table 1.
Multiply the rate by 95% to obtain the usable bandwidth in an STS-1 (reduction due to overhead bytes).

STS

Optical Carrier

Optical Carrier Rate (Mb/s)

STS-1

OC-1

51.840

STS-3

OC-3

155.520

STS-12

OC-12

622.080

STS-24

OC-24

1244.160

STS-48

OC-48

2488.320

STS-192

OC-192

9953.280

STS-768

OC-768

39812.120

STS-3072

OC-3072

159252.240


Table 1 SONET/Synchronous Digital Hierarchy (SDH)

One OC-48 can carry approximately 2.5 channels of 1 Gb/s traffic, as shown in Table 1. To achieve higher data rates for client connections, multiple STS-1s are byte-interleaved to create an STS-N. SONET defines this as byte-interleaving three STS-1s into an STS-3, and subsequently interleaving STS-3s.
By definition, each STS is still visible and available for ADD/DROP multiplexing in SONET, although most SAN requirements can be met with less complex point-to-point connections. The addition of DWDM can even further consolidate multiple SONET connections (OC-48), while also providing distance extension.

GbE
Gigabit Ethernet (GbE) is a terminology describing an array of technologies involved in the transmission of Ethernet packets at the rate of 1024 megabits (Mb/s) or 1 gigabit per second. Gigabit Ethernet is specifically designed to surpass the traditional 10/100 Mb/s link speeds. GbE is defined by the IEEE publication 802.3z, which was standardized in June, 1998. This is a physical layer standard following elements of the ANSI Fibre Channel’s physical layer. This standard is one of many additions to the original Ethernet standard (802.3 - Ethernet Frame) published in 1985 by the IEEE organization. The following are nomenclature and characteristics of GbE.
◆ 1000Base-SX is defined as a fiber-optic Gigabit Ethernet standard encompassing the use of multi-mode (50 or 62.5 micron) fiber with 850 nanometer wavelengths. Distances of over 500 meters can be achieved.
◆ 1000Base-Lx is defined as a fiber-optic Gigabit Ethernet standard encompassing the use of single-mode (9 micron) fiber with 1310 nanometer wavelengths. Distances of 10 km or more can be achieved.
◆ Copper coaxial cabling, multi-mode fiber-optic cabling (50 and 62.5 micron) and single-mode (9 micron) cabling are available choices for the 802.3z standard.
◆ GbE is mainly used in distance extension products as the transport layer for protocol such as TCP/IP. However, in some cases the product is based on a vendor-unique protocol.
◆ Distance products using GbE may offer features such as compression, write acceleration, and buffer credit spoofing

TCP/IP
As discussed in, ”TCP/IP technology overview” the Transmission Control Protocol (TCP) is a connection-oriented transport protocol that guarantees reliable in-order delivery of a stream of bytes between the endpoints of a connection. TCP achieves this by assigning each byte of data a unique sequence number, maintaining timers, acknowledging received data through the use of acknowledgements (ACKs), and retransmission of data if necessary. Once a connection is established between the endpoints data can be transferred. The data stream that passes across the connection is considered a single sequence of eight-bit bytes, each of which is given a sequence number.
For information on the following, refer to, ”TCP/IP technology overview”:
◆ “TCP terminology
◆ “TCP error recovery
◆ “Network congestion
◆ “Internet Protocol security (IPsec)
◆ “Tunneling and IPsec