10 Gb oversubscribed port for Dark Fiber ring

Hi, I understand that for an 8-port 10Gb module (6509) ports 1,2,5 & 6 are the only performance 10 Gb ports (true 20Gb at full duplex) Does this mean I will not get the 10Gb throughput for another port, I'm debating whether or not it is worth moving my 10Gb optical circuit (dark fiber) on a port #4 to a "performance" port such as 1,2,5 or 6. Thanks to all for your input.

The 6708-10G card has 2 connections to the switch fabric
Fabric Channel #1: Ports 2, 3, 6, 8
Fabric Channel #2: Ports 1, 4, 5, 7
As far as i understand it that means that if you want 10Gbps on the port you can only use 2 ports from each fabric channel.
Cisco have provided a command to automatically shutdown ports 3,4,7,8 so that you are not oversubscribed on this module. But i don't think you have to use 1,2,5 & 6. You could use port 4 as long as you only used one other port from that fabric channel group ie. 1,5 or 7 and you would still get the full 10Gbps throughput.
Note i haven't actually tested the above but it is based on the way that all the 6500 fabric enabled modules work.
Jon

Similar Messages

Connecting two CRS-4 over 110Km lenght dark-fiber

I wonder if I want connect two CRS-4 over 110Km lenght and get 40G of througput whitout ONS 15454 MSTP DWDM infrastructure.
I am looking for lowering the cost for a two cities point-to-point carrier class connection with high throughput.
TIA

Thank you for the reply.
As long as I can understand for it, it is possible to interconnect two CRS-4 by mean a dark-fiber without a DWDM transport system, if the distance and fiber parameters limits are meeted.
In the 1-Port OC-768C/STM-256C Tunable WDMPOS datasheet I can read that the targeted distance is 1000km for DWDM Line Interface, and 80 or 100Km for Single-channel optical link (without DWDM).
This would mean that a DWDM platform is needed to get the 1000km length?
What about the throughput when it is connected in single-chanel mode?
http://www.cisco.com/en/US/prod/collateral/routers/ps5763/product_data_sheet0900aecd80395bbe.html
I appreciate so much your attention.
Gustavo Espitia

ATT Fiber Ring 50MB line - no negotiating

Hello,
Firstly the situation we have here is that we are on the outskirts of an ATT Fiber Ring and we have servers currently hosted at an ATT Datacenter. They stated we could hop on that fiber box with their help and get a 50MB link to our cage in the datacenter...great news huh? They also stated that they would give us an Ethernet handoff...well they didn't...fiber instead. So we went out and bought 2 3560G with GBICs, one for each end. We tested the switches inside the office with a fiber cable and everything was fine.
So then we hooked up these switches to the Cisco ONS 15327 and ATT repeatedly told us that the circuit was fine...no problem...great. Still no data...they said just hook them up and ping...nothing else to configure. We did this and nothing happened. Also when doing a "show int gig0/49" we get line down and no lights on the outside of the GBIC on the 3560G. But when we flip the fiber cables the wrong way the 15327 seems to respond so there doesn't seem to be anything physically wrong with our setup.
I'm not terribly great with these things but ATT is not helping us at all. They said they have no Cisco techs that could help us...don't know why...seems pretty simple to me...we must be missing some sort of protocol or something for the handshake that isn't happening. I have included a copy of the configs maybe someone out there can at least shine some light before we tell ATT to shove it.
Thank you

The Cisco ONS 15305 transports Ethernet and TDM traffic inside an SDH frame for metropolitan-area network (MAN or metro) applications. The Cisco ONS 15305 can be used as customer premises equipment (CPE) for medium- and large-sized businesses - collecting voice and data traffic from each site for switched voiced and data services, Layer 1 and Layer 2 VPNs, and Internet access - for transport over an STM-1, STM-4, or STM-16 SDH uplink to the central office. The Cisco ONS 15305 can also be deployed as a central office aggregator; consolidating STM-1 access links from Cisco ONS 15302 Multiservice Provisioning Platforms or other Cisco ONS 15305 CPE, as well as directly connected customer TDM or data traffic from electrical interfaces

It's happened again! Problem with replacement screen (mysterious dark O ring in the centre)

I had my screen replaced for a dark O ring appearing in the centre about half a year ago. It must be caused by the apple logo on the reverse.
It's now starting to happen again and my 3 year applecare has run out
What can I do? Does a new screen come with a guarantee of any kind?
Serial is W89471B364B

Welcome to the forum!
This has been a fairly common problem with Z series units, and the cure is LCD replacement.
Brand new LCDs change hands for under $200 on eBay. If you are handy and careful enough, replacing the LCD isn't that much of a problem.
Hope this helps and good luck.
Cheers,
George
In daily use: R60F, R500F, T61, T410
Collecting dust: T60
Enjoying retirement: A31p, T42p,
Non-ThinkPads: Panasonic CF-31 & CF-52, HP 8760W
Starting Thursday, 08/14/2014 I'll be away from the forums until further notice. Please do NOT send private messages since I won't be able to read them. Thank you.

What Gear do i need to enable WDM (DWDM) on a Dark Fiber?

We have Dark Fiber between 2 DCs and would like to enable WDM on it. What Cisco Optical Gear do i need for this? We would also link to enable Encryption on this link as well. TIA

You can do "simple" CWDM with the Cisco WDM series of CWDM Passive devices. Those can support 4 or 8 channels.
They depend on you "feeding" them signals from special transceiver models that modulate the light on a wavelength compliant with the ITU-50GHz channels listed in the DWDM SFP+ data sheet Leo mentioned and linked to.
The compatibility matrix tells you what Cisco gear is compatible with those transceivers. If you're using other 3rd party gear, they have equivalent support matrices. The bottom line is you need to match wavelengths to what the mux is expecting.
If you move up to the ONS 15454 line you have more flexibility - transponder cards etc that will transform standard SFP+ (1550 nm 10 Gbps signalling on fiber) to the ITU channels etc. You're talking serious money there though - hundreds of thousands of US$. Cisco will be happy to send a sales engineer over and build a nice detailed bill of materials if that's in your budget.

Extending SAN&LAN over dark fiber (DWDM)?

Hi, sorry for my English, but I hope you understand me, and I understand your answers:)
We are in the process of adding new data center (DC2), which will be geographically located about 20 km from our existing data center (DC1). Between them laid dark fiber (two fibers). We need a solution for the LAN(layer 2) and SAN(native FC) extension ...
Solution is to see me use two passive EWDM-MUX8 at both ends of dark fiber with DWDM transceivers on the equipment in the DC1 & DC2.
For Ethernet, we propose to use Cisco 10GBASE DWDM X2 Modules (p/n:DWDM-X2-xx.xx =) - equipment at both ends supports these modules.
For FC we propose to use DWDM SFP Modules for Cisco ONS Family (p/n:ONS-SC-4G-xx.x =).
SAN equipment at both ends - Cisco MDS 9124 (p/n:DS-C9124AP-K9).
But I'm not entirely sure of the correctness of this decision, especially in SAN extension.
Is this workable solution?
Does the Cisco MDS 9124 supports this transceivers (ONS-SC-4G-xx.x=)?
There are other potential pitfalls are not visible to me?
In advance, thanks for the advice.

For ethernet, your decision is correct.
For SAN, things get a little dicey. The MDS 9124 doesn't support the ONS transcievers, and the FC 4G LH transciever is limited in distance to about 10 km.
If it were me, I would do things a little differently. I would put a Nexus 5000 series at each end of your dark fiber, with a 20 GB etherchannel trunk between them. Each nexus would then be connected to your data center ethernet network via 10 gig ethernet, and to your MDS's via 2/4/8 gig FC. The nexus could then be set up such that FC traffic destined for the other end will be encapsulated in ethernet, sent over the link, unencapsulated at the other end, and sent to the appropriate FC host.
It's relatively easy, simple to operate, and no mucking around with DWDM/CWDM, which can be a very complex animal.

MACSec over dark fiber - transceiver module option?

Dear all,
Want to order two WS-C3750X-24T-S with service modules C3KX-SM-10G in order to implement MACSec between two datacenters using dark fiber (one fiber). Distance between datacenters is about 17 Km.
Can I use GLC-EX-SMD module on both location?
Thanks in advance,

Hi Marvin,
Thanks for the response,
Actually I'm little bit confused because ISP fiber engineer explain to me that since there is single fiber optic between two datacenters he suggest me to use GLC-BX40-D-I in one side and GLC-BX40-U-I in other side. This is because GLC-BX40-D-I 1550nm for TX and 1310 for RX and viceversa for GLC-BX40-U-I. In other hand GLC-EX-SMD operates only on range 1310nm and this module will be installed in both sides.
Based on cisco docs GLC-BX40-D-I and GLC-BX40-U-I not supported in 3750X.
I just want to be sure if it will work for sure in my environment?
BR
Enis

How to setup if both node MXP_MR_10DME_C GE down when dark fiber LOS

Hi,
I have both ONS15454 have MXP_MR_10DME_C.
When dark fiber LOS both node MXP GE port status color are green and brown randomly.
May I setup when dark fiber LOS then both node GE status down? How?
Thanks,
Adam

Multiple WAN site redundancy design review (dark fiber, p2p, DMVPN)

I'm re-designing a couple of wan sites. I'm using EIGRP over both some leased dark fiber and p2p provider connections. The attached (pdf) physical topology says it all, I'm thinking of using ip sla to track and inject routes over prefered connections, but really just looking for feed back if someone is interested in taking a look.
I've bought 2 2951's with es3g-16-p modules so I can build svi's and do hsrp between the paths, building redundancy between the 3 available paths back to our enterprise core (1Gbps, 40Mbps, 50Mbps).
multiple vlans at both sites...
e.g.: (wan site1 (vlan 10-15), want site2 (vlan 16-20))
Thoughts and thanks?

hi there
not sure why you need to use DMVPN if it all internal same internal network unless you need to have all the traffic between sites to be encrypted
anyway in general i would say of use the direct link to reach the directly connected networks per site
example using site one 100M link to reach DC and WAN
and use site2 50M local link to reach WAN as primary path and use the site1-site2 fibre to reach DC as primary path for site2 this could archive a good load sharing and reduce the load on the link between site1 and site2
IP SLA in a topology like your for sure can very helpful to improve failover time and make the routing more topology aware
hope this helps

Testing New dark fiber installation

What is the best way and/or best software for testing a new installation of dark fiber, not only the physical but also test for L2 errors etc, the fiber connects / terminates at two 6500's and will be configuring a /30 network as to not disrupt anything else in the network, thanks in advance.

In addition to OTDR testing of the fiber, once you plug it in and assign IP addresses from the /30 subnet, you could run IP Service Level Agreement (SLA) performance tests between the two 6500s.
Set one end up to be the "sender" and the other end to be the "responder". Configure a test or simulation that generates packet traffic between them. Sender sends to responder, which in turn returns it to sender. If you have NTP time synchronization configured on the 6500 switches, IP SLA can measure not only packet loss (indicating a questionable or saturated link) and in which direction the loss occurred; it can also report on the variability in delivery timing between packets, and lots more.
Depending on the age of your IOS, it may be termed IP SAA (for Service Assurance Agent) instead of IP SLA.
See the Cisco White Paper "Cisco IOS IP Service Level Agreements" at the following link for more details:
http://www.cisco.com/en/US/technologies/tk648/tk362/tk920/technologies_white_paper0900aecd8017f8c9.html
Oh, and I forgot to mention the best part: it's free.

Backup interface on fiber ring

We recently installed a 100Meg MPLS fiber ring into our office, and we were given redundant connections (north loop and south loop).
The North loop is the primary interface, and will be used unless the fiber is cut.
I need to configure our router to automatically switch to the south loop in the event that the north loop goes down.
Here is the config I have so far, I just want to make sure I'm not missing anything.
interface GigabitEthernet0/0/0
description 100MB MPLS North Loop
backup delay 30 never
backup interface GigabitEthernet0/0/1
ip address X.X.X.X 255.255.255.252
media-type sfp
negotiation auto
interface GigabitEthernet0/0/1
description 100MB MPLS backup South Loop
ip address X.X.X.X 255.255.255.252
media-type sfp
negotiation auto
Am I correct in thinking that, should the North loop on Gi0/0/0 go down, it will automatically switch to Gi0/0/1 after 30 seconds of down time?
If so, how do I go about switching back to the north loop when it comes back online?

Hi,
I have see this feature configure on few ISP. For this feature to work, interface needs to go down then only RP can send UP event to standby link and make it up.
You can also go through below discussion on backup interface
https://supportforums.cisco.com/discussion/10921951/how-set-backup-delay-serial-interface-ethernet-interface
regarding backup delay , i am not sure why you want to put delay of 30 sec before standby interface becomes active and then deactivating delay is infinite, means it will always remain active even when primary interface comes up.
CCO document on backup delay
http://www.cisco.com/c/en/us/td/docs/ios/dial/command/reference/dia-cr-book/dia_a1.html#wp1011975
backup delay
To define how much time should elapse before a secondary line status changes after a primary line status has changed, use the backup delaycommand in interface configuration mode. To return to the default so that as soon as the primary fails, the secondary is immediately brought up without delay, use the no form of this command.
backup delay {enable-delay-period | never} {disable-delay-period | never}
no backup delay {enable-delay-period | never} {disable-delay-period | never}
Syntax Description
enable-delay-period
Number of seconds that elapse after the primary line goes down before the Cisco IOS software activates the secondary line.
disable-delay-period
Number of seconds that elapse after the primary line comes up before the Cisco IOS software deactivates the secondary line.
never
Secondary line is never activated or deactivated.
--Pls rate helpful posts--
Regards,
Akash

I have a Mac OSX version 10.75 with just one Thunderbolt port. and it has been my Thunderbolt port to connect with Blackmagic wear my intensity. and I no longer can use the port for mini-DVI adapter to connect with me. I do not want to ask any other way f

i have a Mac OSX version 10.75 with just one Thunderbolt port. and it has been my Thunderbolt port to connect with Blackmagic wear my intensity. and I no longer can use the port for mini-DVI adapter to connect with me. I do not want to ask any other way for me to use to use my monitor. I monitor LG FLATRON E2041 brand .. PLEASE Helpp ME

i have a Mac OSX version 10.75 with just one Thunderbolt port. and it has been my Thunderbolt port to connect with Blackmagic wear my intensity. and I no longer can use the port for mini-DVI adapter to connect with me. I do not want to ask any other way for me to use to use my monitor. I monitor LG FLATRON E2041 brand .. PLEASE Helpp ME

Open ports for all in LAN

Hi, a few days ago I bought a wireless router WRT160n. I want to ask how to open some port for all in LAN(3 clients). For example all in LAN have PeerToPeer application for torrents. I want to open port for example 20202 for all. Now I open port from menu Applications & Gaming->Single Port Forwarding, but I must set port for each user IP address.
Can somebody tell me how to open port for all in LAN without to config for each computer?
Thanks in advance.

Hi gv. I read more about UPnP and the WRT160n User Guide. In section Administration>managment int wrote that UPnP is Enabled by default in my router it is corect. I Setup mu PeerToPeer(eMule) TCP/UDP ports to 20202 and check option "Use UPnP to setup ports". I test and close this port on my router configuration for my computer on "Single Port Forwarding", but in eMule the port is still block. Can you explain why it did not work. For UPnP it says that if Enable it allow users with Windows ME and XP automatically to gonfigure Router ports
Thanks in advance!
Best Regards.

Can there be more than one port for the same logical system

Can we have more than one TRFC port for tranferring data from one logical system to another?
The requirement is to transfer same basic idoc type from one logical system to another through two different ports. Is it possible?
For eg. some specific data (country specific) for the same basic idoc type can be sent through a separate port and the other data through a different port inorder to distinguish between the two.

Hi Shital,
I don't think so. Because of that what i told you. There would be a dirty trick (back routing) to solve your issue. But its quite difficult and not standard. You create two http receiver services and two http sender services. You route your message depending on payload to the one or the other http receiver channel. The receiver is in both cases: THE XI! Now you have a second message, which you can route to IDoc adapter. But in this case you have different sender and in the receiver determination you can put the sender, so you will be able to use different communication channnels and so different ports / destinations.
Advantage: Your task would run, no performance problem (http: very good, asynchron)
Handicap: You see every message double in the monitoring, the scenario is not easy to understand and it is a big task 4 other people to change it.
Regards,
Udo

How do I create Labview VISA ports for *individual* GPIB instruments using Prologix USB GPIB controller?

Hello,
I'm trying to use a Prologix USB GPIB controller to control GPIB
instruments, and I would like to have a virtual serial (VISA) port for
*each instrument*, as is the case with a normal GPIB controller with a
standard NI driver. However this is not what the Prologix driver
provides -- it provides a single VISA virtual serial port for the
entire controller. To address the instrument with GPIB address 11,
you first send "++addr 11" to the serial port, and then you're talking
to instrument 11. However, this means I have to change all old
Labview programs.
Is it possible to create a "wrapper" function of some kind that will
define a virtual serial (VISA) port for each *instrument* on the
controller? For example, to talk to GPIB instrument 11, call it
ASRL3::11::INSTR, each time it is written to it would have to write to
the virtual serial port of the controller, say ASRL3::INSTR, first "+
+addr 11" and then the command that is sent to it.
A clearer explanation of the difference (i.e. incompatibility), and of
my objective:
1) A normal GPIB controller with NI driver: I go to the NI
Measurement & Instrumentation Panel, under GPIB, and Scan for
Instruments; all the live instruments show up; subsequently when I
want to use Labview programs that use VISA ports, the VISA drop boxes
allow me to choose a different port for each instrument, e.g.
"GPIB0::11::INSTR", "GPIB0::12::INSTR" would be instruments at
addresses GPIB 11 and GPIB 12.
2) The Prologix GPIB controller that plugs into a USB port: In
Labview you get a *single* VISA virtual serial port, ASRL3::INSTR, for
the entire GPIB0 controller. Therefore to address GPIB instrument 11,
you write "++addr 11" to the virtual serial port ASRL3::INSTR, and
then you are communicating with device 11, so you can write and read
ASRL3::INSTR to talk to that device. Then to talk to device GPIB 12,
you write "++addr 12" to the same VISA port, and then you are talking
to that device. The problem is that this requires recoding all
Labview code, whereas I would like to be able to use the same program
either with a normal or with a Prologix GPIB controller. Therefore, I
would like to create code that scans the controller for all GPIB
attached devices and creates VISA ports for all. Such ports, when
written to, would have to first write "++addr DEVICENUM" to
ASRL3::INSTR (i.e. the port of the GPIB-USB controller) where
DEVICENUM is the GPIB address of the instrument corresponding to that
port, and then would have to do a write or read or whatever function
is being done on that instrument VISA port.
I haven't figured out if it is possible to do this easily. Help and
pointers on where to look for hints would be much appreciated. Many
thanks!
Milos

My first impression is that if you don't want to make any changes at all to existing programs is that the wrapper you need is one around VISA. You would need to intercept all of the calls into the NI VISA driver. If you create your own visa32.dll and in there, change the addressing and then call the real VISA driver, you might (repeat, might) get something to work. If this would even work, you still might find that you have to make significant changes anyway. The serial connection is going to be considerably slower, and interface specific functions such as service request handling, bus triggering of multiple instruments, etc., would be difficult to impossible. This would be a lot of work, imho, to just save a couple of hundred dollars over a real GPIB controller. I've seen this Prologix device before and have even used NI's RS-232->GPIB controller. The Prologix intended use to me seems to me more for a hobbyist or very casual user. Of course, I'm used to having multiple GPIB instruments worth 10s/100s of thousands of dollars and the cost of an fully compliant GPIB controller is just lost in the noise.

10 Gb oversubscribed port for Dark Fiber ring

Similar Messages

Maybe you are looking for