Backplane switching

I have got a 4506 with Sup V 10GE module, Can anybody say me how to monitor the backplane switching utilization.

There is no such command. The IOS running in my switch is 12.2(25)EWA (cat4000-i9s-mz.122-25.EWA.bin).

Similar Messages

Supervisory engine, backplane switching,line cards

Hello,
i have a couple of questions and they are thus:
1.Is the line card a module or are they different parts of the switch. What is the function of the line card ?
2.From my understanding the supervisory engine is like the brain of the device. It controls all or most of the software related functionality of the device. Am I correct?
3.What is the difference between the SUP720 and SUP32?
4.Is the CatOS solely for layer 2 switching while the IOS is for routing?
5. when we refer to backplane does that mean how fast a switch switches packets ?
Best Regards,
DJ.

Hello,
Please find below answers inline,
1.Is the line card a module or are they different parts of the switch. What is the function of the line card ?
Yes, Linecards/Modules are different part of the switch. You can insert multiple cards into the Chassis. You can also have service moduels like Wism/FWSM/ACE
depends upon the business needs.
2.From my understanding the supervisory engine is like the brain of the device. It controls all or most of the software related functionality of the device. Am I correct?
Yes, Supervisor is the brain of the switch controls each function. A typical Supervisor consists of PFC which has TCAM L2/L3 Engines and RP/SP Processor and A replication engine (Hyperion) for Multicast. TCAM is the area where the RP/SP programmes the switch. There are multiple things that get programmed in TCAM like QoS/ACL/Netflow etc.
They have their respective hardware entries in the TCAM. If the entry is not found that is when the packet gets punted to CPU. If the linecard has a DFC on it then it can download the information from L2/L3 Engines from TCAM and switching gets fast as everytime the lookup happens in the Linecard instead of the Supervisor.
3.What is the difference between the SUP720 and SUP32?
SUp 32 : 32Gb backplane supporting hardware accelerated Layer 2 and 3, QoS and Security policies up to 15Mpps…
Sup 720: 720Gb backplane supporting hardware accelerated Layer 2 and 3, QoS and Security policies up to 400Mpps…
4.Is the CatOS solely for layer 2 switching while the IOS is for routing?
If you have a hybrid switch with CATOS/MSFC(IOS) CATos represents the switching part and MSFC represents the Routing part of the network.
5. when we refer to backplane does that mean how fast a switch switches packets ?
Backplane is used for communication between SUP - LC's.The area divided into three subsections,
1) D-bus - Used for data flow
2) R-bus - Used for result lookups
3) EOBC - Control communications like IPC/ScP which is helps the Supervisor to maintain the other line cards.
Depends on the line card the backplane can work in three different modes,
1) Flowthrough
Between non fabric modules and between a non fabric and a fabric enabled linecard
Throughput – 15 Mpps (@ 64 byte frames)
Bandwidth – 16 Gbps of bandwidth shared throughout
Data Bus frame size is variable; min of 4 cycles (64B Data) on the DBus for every frame +1 wait cycle
2) compact
When only ALL fabric enabled linecards in a chassis
Throughput – 30 Mpps (@ any frame size)
Bandwidth – 8 G CEF256; 20 G/channel CEF720
Data Bus frame size is constant (compact header); 2 cycles (32 B Data) on the DBus for every frame + no wait cycle
3) truncated
Between fabric linecards when a non fabric linecard is in the chassis.
Throughput – 15 Mpps (@ 64 byte frames); independent of frame size for CEF256 and CEF720
Bandwidth – 16 G shared for classic; 8 G per CEF256; 20 G/channel CEF720.
Data Bus frame size is variable; min of 4 cycles (64 Bytes Data) on the Data Bus for every frame.
Thanks,
Ricky Micky
*Rate Useful posts

CSCuh36583 - ISRG2 - DSP Crash due to Backplane Switch Lockup During T1 Errors

Hi, I got same problem with IOS Version 15.1(4)M7
Is it a bug to this version too?

radionecrotic wrote:
Has this always been the case for you or did you update recently and it started happening? There appears to be some sort of conflict with some machines and new changes to the kernel to allow quicker mode setting. See http://bbs.archlinux.org/viewtopic.php? … 47#p659347. I had to add !radeon to my MODULES list in /etc/rc.conf to get X to even work at all. I don't run anything that uses OpenGL so I don't know if it'll address your issue.
You may also want to take a look at http://bbs.archlinux.org/viewtopic.php? … 20#p662920 and add intel_agp radeon (that's two modules, in that order) to your MODULES list.
Well, I was having the graphical corruption issue in that first thread until I added the nomodeset boot flag. As for the modesetting, I think it directly conflicts with my boot flag, and without it, I get corruption. =/

Maximum 29XX switches on a 6500 series

Hello,
is there a recommended maximum of user ports i can connect directly to a 6500 series core switch?
So I have a situation with about 80 24-ports switches (in stacks of 2 or 3) connected with fiber to the core switch. Now one of the network managers states we need a distribution layer cause performance in the 6500 will drop. Reason for this is that the 29XX series switches do not support layer 3.
Is this correct? And does the added layer 3 functionality in a (extra) distribution layer increase performance in the core? Even though almost all traffic is user - server traffic and servers are connected to core?
Anyone know? THanks!

Hi Friend,
Which sup engine do you have and is there any switch fabric module you are running. Cat6500' are cross-fabric chassis which gives 32GBPS backplane switch capacity bedefualt. If you have SFM (switch fabric module) installed then it will give 256 GBPS of backplane capacity.
If you are running Sup720, you have 720 GBPS backplane capacity and you will see a greater performance by cat65K.
If you dont have SFM or sup 720, then yes with direct access switches uplink to Core might slow down the performance of the core switch.
If that's the case I would suggest to got for hierarchical design and add another layer of the architecture i.e distribution layer.Now what you can do with this additional layer, is that you can uplink your access layer to it, have your inter-vlan routing done on thius layer, have to access control policy deployed at this layer.
This addtion layer and the above functionalities will off load the processing of traffic at core and core will be efficient enough to process the traffic as fast as it can. This is a abosultely correct approach.
Please, write back for any qurstions.
HTH, Please rate if it does.
-amit singh

P2p routing

I have 7 flexrio cards 5 7965 and 2 7961 and am heavily using P2P streaming to transfer data between them. The chassis is a PXIe-1065. Having enable all the P2P connections is there anyway of knowing which of the links go by the backplane switches and which have to use the host PC?

The P2P connections should all take a direct (or the most direct) route to each other:
From looking at this diagram, I can say that P2P connections on boards 9-14 should route through the chassis' PCIe Switch and should therefore communicate at expected P2P speeds.
A P2P connection to a card in slot 7 or 8 will be entirely dependent on the architecture of the controller. In addition to this, if multiple streams are transferred through the same switch then the 800 MB/s maximum transfer speed will be distributed (Two simultaneous transfers using four cards should mean that you experience a transfer rate slightly below 400 MB/s). Does this fit with the pattern of data rates you are experiencing?
Regards,
Ben Clark

Differences between 6500-E and 6500 chassis?

Can someone please provide a link to a Cisco url/document describing the differences between the 6500-E and 6500 series chassis? I've heard references to an increase in backplane/switch fabric throughput and also an increase in number of POE devices supported but nothing that can be verified on the Cisco site. Thank you in advance.

Long,
I have not seen a specific document that compares the 6500 and 6500-E Chassis.
The only major difference between the 6500 and 6500-E Chassis is the 6500-E series switches offer enhancements to scale beyond the 4000W power supply. The 6500-E provides higher density PoE deployments that require more power than the original 6500 chassis can deliver. My undertanding is, there is no difference in backplane throughput between the 6500 and 6500-E Chassis.
Cisco Catalyst 6500 and 6500-E Series Switch Data Sheet
http://www.cisco.com/en/US/products/hw/switches/ps708/products_data_sheet09186a00800ff916.html
Let me know if it helps.
Regards,
Arul

How i can calculate the backplane speed & throughput of cisco 48 1G 2960S switch?

How i can calculate the backplane speed & throughput of cisco 48 1G 2960S switch?

Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
Calculate? Calculate for wirespeed/line-rate? If the latter, take all the port bandwidths, and assuming they are duplex, double for necessary fabric bandwidth. I.e. 48 gig ports would need a 96 Gbps fabric. Take all your port bandwidths, and allow 1.448 Mpps per gig (for minimum size Ethernet packets), i.e. 48 gig ports would need 69.5 Mpps. Once you have required fabric bandwidth and PPS, you can compare to vendor's specs.

Switching Backplane

Hi
I want to know about the method through which we can calculate the Total Backplane throughput of a switch like 6500 series.
If anybody know please reply to me.
Thanks & Regards
Rahul Chhabra

Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
3)
there is no way you can increase the backplane capacity other than replacing the chassis.
Well, yes and no.
The 6500 series "backplane" capacity depends on the chassis (pre -E offer up to 40 Gbps per slot, -E offer up to 80 Gbps per slot), the supervisor (sup1 and 2, natively, only support the 32 Gbps bus), the supervisor and fabric card (sup2 with SFM provides 256 Gbps fabric, sup720 provides 720 Gbps fabric, sup2T provides 2 Tbps fabric), the line card (from 32 Gbps bus to 80 Gbps fabric), and even the combination of type of line cards (bus only cards preclude full fabric operation mode, even for installed fabric capable cards) and/or how chassis is configured (combination bus CEF256 line cards offer a choice of which they use to transmit to other cards). Also with fabric cards, it's possible DFCs are enabled to fully exploit all the fabric bandwidth (i.e. a sup, alone, PPS capacity can bottleneck packet forwarding).
From a configuration only view, pretty much only the bus/CEF256 cards transmission choice can impact performance, otherwise you'll likely need to change hardware, but depending on your needs, just adding a DFC or replacing bus only cards can make a huge difference (although you may need to really load the chassis with traffic to notice). I.e. you may not need to replace the chassis or all your cards. (BTW, to fully exploit a sup2T, you need either a 6513-E or a 6807, the latter probably a better choice for a new chassis.)

Auto switch to BNC ref-in to replace backplane 10MHz if detected clock signal there with PXI-6608 on Slot2

Hi,
I m using chassis PXI-1045. The priority of backplane frequency is Slot 2 => Ref-In => Origin Backplane 10MHz.
But I wanted the backplane frequency to be in the priority order of Ref-In => Slot 2 => Origin Backplane 10MHz.
So,
It is possible to change the priority ? maybe by edit the config file somewhere else ?
If the priority is uneditable, is there any function/VI to disable the Slot 2 PXI-6608 from replace the backplane 10MHz ? (from LabVIEW interface) Let say disable the device without access to MAX ?
Thanks a loottt
BR,
engwei

Hi Eng Wei,
I've checked with R&D, and the priority of the clock selection is embedded in the firmware on the backplane. So unfortunately it is not possible to change the priority via software.
As far as the DAQmx functions are concerned, I did a check but was also not able to find any function that can disable the NI 6608 clock from being used to override the PXI backplane clock.
Is it a must that the NI 6608 be placed in slot 2, if it is not used to drive the backplane?
Best regards,
Victor
NI ASEAN

Problem with FWSM and L3 interface in same switch

I have two 6513s with an 802.1q trunk connecting them. Each switch has redundant Sup720s running in Native mode, IOS ver 12.2(18)SXF (they were initially running SXD3). A FWSM (ver 2.3(3), routed mode, single context) is in each switch, setup in failover mode.
I can not get a PC, in a vlan that has the layer 3 interface defined on the switch with the active FWSM in it, to communicate with devices "behind" the FWSM. If I move the layer 3 configuration for that vlan to the other 6513, everything works fine.
The MSFCs are on the inside of the firewall, they have a layer 3 interface configured in the same vlan as the FWSM "inside" interface. Several "same security level" interfaces are defined on the FWSM and used to protect server farms. I am using OSPF on the MSFCs and FWSM and the routing table is correct.
The FWSM builds connections for attempts made by the PC with the layer 3 interface defined on the same switch as the active FWSM just fine, so this is not a FWSM ACL problem.
A ping of the FWSM "inside" interface from a PC with the layer 3 interface defined on the same switch as the active FWSM fails, even though debug icmp trace on the FWSM shows the request and the response. A packet capture, using the NAM-2, shows only the request packets. I have captured on the common vlan and the FWSM backplane port channel interface.
Just to add to the confusion, if I capture in the same places, but do the ping from a PC that is in a vlan with the layer 3 interface defined in the 6513 that does not contain the active FWSM, which works fine, I see the request and reply on the common vlan capture, but only the request on the port channel capture.
This problem has been there from the beginning of this implementation and has not changed with IOS and FWSM software upgrades. I have experienced this with any and all vlans that I tried to define the layer 3 interface for on the switch with the active FWSM. I have MLS turned on.
If anyone else has experienced this and solved it, or knows what is going on, I would appreciate any insight.
Thanks.
Keith

I will have to get setup to record more data, but I do know the FWSM showed a ping request and a ping reply at the "inside" interface.
I believe my problem is related to the IOS command "firewall multiple-vlan-interfaces" which I put in place to allow IPX traffic to be brought around the FWSM. The little documentation that there is for this command, states that policy routing may need to be implemented to prevent ip packets from going around the firewall. I do not have any policy routing in place.
I also do not have any active layer three interfaces defined for any of the vlans assigned to the firewall except the "inside" interface. So my resoning was that I did not need to be concerned about ip packets having a way around the FWSM. My suspicion is that this command and the fact that I have mls on is causing some type of a problem which results in the packet being "lost" when it needs to be going through the MSFC in the switch with the active FWSM to get to the PC. Hopefully that makes some sense.
Do you have any idea where better documention on using the "firewall multiple-vlan-interfaces" may be, or a better explanation of all that is happening inside the switch when that command is used?
Thanks.

Branch office setup with L3 switch and router with IOS security

Hello,
I am in the process of putting together a small branch office network and I am in need of some design advise. The network will support about 10-15 workstations/phones, 3-4 printers, and 4-5 servers. In addition we will eventually have up to 25-30 remote users connecting to the servers via remote access VPN, and there will also be 2-3 site-to-site IPSec tunnels to reach other branches.
I have a 2911 (security bundle) router and 3560 IP Base L3 switch to work with. I have attached a basic diagram of my topology. My initial design plan for the network was to setup separate VLANs for workstation, phone, printer, and server traffic. The 3560 would then be setup with SVIs to perform routing between VLANs. The port between the router and switch would be setup as a routed port, and static routes would be applied on the switch and router as necessary. The thought behind this was that I'd be utilizing the switch backplane for VLAN routing instead instead of doing router-on-a-stick.
Since there is no firewall between the switch and router my plan was to setup IOS firewalling on the router. From what I am reading ZBF is my best option for this. What I was hoping for was a way to set custom policies for each VLAN, but it seems that zones are applied per interface. Since the interface between the router and switch is a routed interface, not a trunk/subinterface(s), it doesn't seem like there would be a way for me to use ZBF to control traffic on different VLANs. From what I am gathering I would have to group all of my internal network into one zone, or I would have to scrap L3 switching all together and do router-on-a-stick if I want to be able to set separate policies for each VLAN. Am I correct in my thinking here?
I guess what I am getting at is that I really don't want to do router-on-a-stick if I have a nice switch backplane to do all of the internal routing. At the same time I obviously need some kind of firewalling done on the router, and since different VLANs have different security requirements the firewalling needs to be fairly granular.
If I am indeed correct in the above thinking what would be the best solution for my scenario? That is, how can I setup this network so that I am utilizing the switch to do L3 routing while also leveraging the firewall capabilities of IOS security?
Any input would be appreciated.
Thanks,
Austin

Thanks for the input.
1. I agree, since I have only three to four printers, they need not be in a separate VLAN. I simply was compartmentalizing VLANs by function when I initially came up with the design.
2. Here's a little more info on the phone situation. The phones are VoIP. The IP PBX is on premise, but they are currently on a completely separate ISP/network. The goal in the future is to converge the data and voice networks and setup PBR/route maps to route voice traffic out the voice ISP and data traffic out the other ISP. This leads up to #3.
3. The reason a router was purchased over a firewall was that ASA's cannot handle routing and dual ISPs very well. PBR is not supported at all on an ASA, and dual ISPs can only be setup in an active/standby state. Also, an ASA Sec+ does not have near the VPN capabilities that the 2911 security does. The ASA Sec+ would support only 25 concurrent IPSec connections while the 2911 security is capable of doing an upwards of 200 IPSec connections.
Your point about moving the SVI's to a firewall to perform filtering between VLANs makes sense, however, wouldn't this be the same thing as creating subinterfaces on a router? In both cases you are moving routing from the switch backplane to the firewall/routing device, which is what I am trying to avoid.

Switch for Distribution Layer

I need your expert opinions for selecting a switch for distribution Layer
My requirement is 24Gig ports @ dist layer which will connect to 3750?s for server farm. I don?t want to negotiate with the bandwidth.
Which switch will be best for this scenario with max Backplane speed

Yes L3 routing will there.
I am considering 3750 for access switch for my server farm, and 4506 for distribution layer with the following port requirement
Catalyst 4500 Gigabit Ethernet Module, 6-Ports (GBIC)
Catalyst 4500 GE Module, Server Switching 18-Ports (GBIC)
Catalyst 4500 Enhanced 48-Port 10/100/1000 Base-T (RJ-45)
Catalyst 4500 Supervisor IV (2 GE), Console (RJ-45)
Most of the gigabit port will be occupied since I have a huge server farm to connect. By using almost all the ports on 4506 will I be able to get the maximum through put??
I have not considered 6500 due to budget. If I consider 6500 will it be a good design to have 6500?s on both core and distribution layer

Pros and Cons of Clustering switches

Can someone tell me the pros and cons of clustering switches? It sound like its just for managing multiple switches or is it more trouble that its worth?
TIA

Hi,
well it has advantages as well as some flaws imho.
Advantage:
- Very scalable at low cost. You can add a lot of models onto a stack, ranging from 12port to 48 port, from 100Base-X Uplink to 10Gig-Uplink, with or without PoE with just the buying of a new switch. In modular switches, you have to buy the chassis and then add modules on it which might not come as divers as standalone-switches.
- A somewhat higher redundancy. Spread the uplinks onto different switches in the stack and you'll have nearly no single point of failure (enviromental errors such as power not considered), whereas modular switches always have the backplane as (very rare) spe... and, if your not rockefeller, is the single supervisor engine.
- You can spread a stack (with 3750) over some distance, at least more than on a modular switch. That gives you enough room to implement some cable-routing facilities inbetween the switches.
Disadvantages:
- surely the need for more power outlets, one for each switch
- management, as stated on previous post (well, it's quite similar to modular switches but not exactly, just have a look at snmp, it's a mess on stacks :-( )
- heat dissipation (not really checked on that). X powersupplies generate more heat than Y, if Y < X...
Mathias

[SOLVED]Planning to Switch - Broadcom "?"

I'm planning on switching from Ubuntu to Arch and am wondering what experiences have been like when dealing with Broadcom wireless cards. In my experience they've been rather a pain to deal with and I just want to find out whether anyone can tell me about their experience getting their cards working under Arch.
Last edited by crisnoh (2008-04-26 14:39:03)

Thanks for letting me know what questions I need to be asking. Since I started using linux, wireless drivers have been my kryptonite!
Looks like b43 is definitely installed. My chipset is: Broadcom Corporation BCM4306 802.11b/g
Here's the relevant info from my "ls -R"
/lib/firmware/b43:
a0g0bsinitvals4.fw a0g1initvals13.fw b0g0initvals4.fw ucode11.fw
a0g0bsinitvals5.fw a0g1initvals5.fw b0g0initvals5.fw ucode13.fw
a0g0initvals4.fw b0g0bsinitvals13.fw lp0bsinitvals13.fw ucode4.fw
a0g0initvals5.fw b0g0bsinitvals4.fw lp0initvals13.fw ucode5.fw
a0g1bsinitvals13.fw b0g0bsinitvals5.fw pcm4.fw
a0g1bsinitvals5.fw b0g0initvals13.fw pcm5.fw
As far as I can tell the module is loading, unless that zero means that it isn't. It should be. I have it listed in the modules section of my rc.conf and have bcm43xx blacklisted:
$ lsmod | grep 'b43\|bcm\ndis'
b43 110752 0
ssb 30340 1 b43
rfkill 5904 3 rfkill_input,b43
pcmcia 33324 2 b43,ssb
mac80211 117644 1 b43
input_polldev 3720 1 b43
pcmcia_core 33304 4 yenta_socket,rsrc_nonstatic,b43,pcmcia
And here is my dmesg result, which seems to say that my wireless card isn't working. Not exactly the 6:00 news:
$ dmesg | grep -C4 'b43\|bcm\|ndis\|wlan0'
PCI: Setting latency timer of device 0000:00:06.1 to 64
ACPI: PCI Interrupt Link [LNK3] enabled at IRQ 17
ACPI: PCI Interrupt 0000:02:02.0[A] -> Link [LNK3] -> GSI 17 (level, low) -> IRQ 20
ssb: Sonics Silicon Backplane found on PCI device 0000:02:02.0
b43-phy0: Broadcom 4306 WLAN found
phy0: Selected rate control algorithm 'simple'
ACPI: AC Adapter [ACAD] (on-line)
ACPI: Battery Slot [BAT1] (battery present)
input: Power Button (FF) as /devices/virtual/input/input2
eth0: link up, 100Mbps, full-duplex, lpa 0x41E1
NET: Registered protocol family 10
lo: Disabled Privacy Extensions
eth0: no IPv6 routers present
input: b43-phy0 as /devices/virtual/input/input9
ADDRCONF(NETDEV_UP): wlan0: link is not ready

How can I power up the 8115 controller without using 1078 momentary contact switch.

I am using an 8115 controller installed in a PXIe-1078 chassis. I am rack mounting the system and I have numerous cables originating at the PXIe-1078 chassis that prevent me from using the front panel momentary switch. Is there a method to have the 8115 boot without using the momentary switch on the PXIe-1078 front panel?

Hello EdH,
Unfortunately there is no way to remotely control the power line of the PXIe-1078. Though the PXIe-1078 has a DIP switch on its backplane this only toggles the control of the Power from the Controller to the Backplane. With this you may be able to mimic a power on sequence but that is unsupported.
Regards,
-Travis E
National Instruments
Applications Engineer

Backplane switching

Similar Messages

Maybe you are looking for