Single or dual 4500e for redundancy

I apologize if this has been asked before, but I haven't been able to find a solid answer when searching.
We're deploying a new datacenter and will have 20 racks of servers, mostly 1u.  Given our bandwidth needs which are reasonably modest we're planning on using dual 2960s switches at the top of the racks and aggregating on a single or dual 4500e using 10gig links.  The 4500e(s) will also have a bunch of 1gb devices such as firewalls, routers, and load balancers connected to them.  Pretty typical network core.    Given the need for 40 10gb links to racks and then another 20-or-so 1gb, the 4500e chassis seems to fit the bill.
My question is, do we go with a single 4507e with full redundancy (dual 10gb cards, one link to each from each rack, dual 1gb card, and dual supervisors) or a pair of either 4506e switches each with their own cards.  With a virtual chassis the 4506e is appealing, but I'm not up to speed on any limitations that might impose.   With IISU and the like does having two physical chassis vs everything in a single actually buy us anything if they're going to be in the same physical proximity anyway?  A big reason to go with the single chassis is space - 11 rack units vs 20.
If we do go with a single chasis, other than Cisco's HA docs on the 4500e, is there any documentation or case studies that I could use to sell the idea to management?  A number of people, especiailly those in favor of a ton of cheap netgear switches, argue that a single chassis is a single point of failure and we should never do it.
Thanks!

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
As you've described, the problem with any single chassis, the chassis itself can become a single point of failure even when loaded with redundant components.  Independent chassis (e.g. VSS, StackWise) provide chassis redundancy although it does raise a question on software redundancy, i.e. multiple chassis are running some form of tightly coupled chassis OS vs. standalone chassis.  Standalone chassis should be, in theory, the most redundant but they too create issues such as load sharing and asymmetrical path issues (e.g. unicast flooding).  If you understand and trust a multi-chassis OS, I personally think it's often the "better" redundancy option.
I don't know if VSS on the 4500s support it, but 6500 VSS supports quad supervisiors, so if you lose a single sup on those you don't lose one of your VSS chassis pair.
Regarding your later post on using a 4500-X VSS pair, yes that might be an idea core for high density 10g.  For twenty 10g ports each, you would need either the 16 port with the optional 8 port module or the 32 port model.
Depending on your remote rack setups, you might also consider 3750-Xs, stacked, in lieu of multiple (individual cabinet) ToR 2960S pairs.  StackWisePlus is a much better stacking technology then FlexStack.  Yes, the 3750-Xs are more expensive, but you might need less if you can bring multiple racks to the same 3750-X stack.  (Depending on how many downstream stacks you actually need, you might also reduce your need for 10g ports on the core.)  Depending on your feature needs, you might even be able to use the LAN Base models which in the later IOSs also support StackPower.
10g is nice, but it's also expensive.  When working with switches within 100m, don't overlook the possibilities of gig copper Etherchannels.  For example compare total cost of 8xgig (copper) build-in ports vs. single 10g (fiber) especially if special module is required.

Similar Messages

  • Single or Dual Processors for Tuxedo and/or weblogic 8.1?

    We run Tools 8.44 and 8.20 and double checking to see if either Weblogic (8.1) or Tuxedo (8.1) would utilize a dual processor environment? We run these on Windows 2003 Server.
    Appreciate inisght.

    Yep its all in the cache, and the L3 makes all the difference.
    I remember when I was shopping for my upgrade, Gigadesign had two different models of CPU's, the (brand new at the time) single and dual 7447's which went up to a dual 1.8 and the older pre-Freescale models (cant remember the model number maybe it was indeed the 7455B??) that maxed out with a dual 1.4..
    I remember reading benchmarks (and quite a bit at XLR8yourmac.com) about the older dual 1.4's being on about the same level as the dual 1.6 and 1.7 7447 cpus.. This was all down to the 1.4 having L3 cache on each cpu. The 7447 had double the L2 like the 7448 but no L3.. I didnt really care much since they were relatively the same benchmark, but I went with the dual 1.7 7447 anyways just because the numbers were higher..
    And funny enough, the CPU never ran stable at 1.7, my Mac would crash and kernel panic left and right. I couldnt get any response back from their tech support (I think it was one guy that ran the whole thing) about adjusting the voltage on it, so it spent its entire lifetime in my G4 as a dual 1.67. Not too big of a difference but I couldve saved some money just flat out buying a 1.6 instead of the 1.7, and wouldve saved even more had I just listened and bought the dual 1.4 with L3...
    John with a G4 - no problem at all hehe! Dont get me wrong I loved my G4 and it served me well for just about 10 years before it 'died' on me.. But after getting the late-2005 dual core G5 man oh man I wish I wouldve done it waaaaaaaaaay sooner!

  • Dual IPsec for redundant purpose (Not Dual ISPs)

      Chapter user (Branch user) usually access data through route #1. When ISP1 is down (route 1), I like to use backup ipsec tunnel (automatically) if possible.
      I saw many articles regarding "One interface with multiple IPsec tunnels"
    http://thelostpackets.blogspot.com/2011/11/multi-point-ipsec-vpn-tunnel.html
      However my requirement is little different because remote IP address is same (10.0.0.0/8) for both IPsec tunnels. VA ---- MPLS ---- MD are connected through MPLS, so remote IP is same.

    May I know what is the difference between
    8GB 1600MHz SODIMM 1.35V    and    8GB 1600MHz SODIMM as you have specified ..  And will Kingston be a good replacement the factory fitted Hynix brand ??
    My laptop is 64 bit Win 8.1 OS

  • Configuring Dual-homed servers for redundancy to 6509s

    I am looking for assitance in properly configuring dual-homed servers for redundancy to two CatOS based 6509s. My search for information on how to do this has been unsuccessful to date, so any help you can provide would be most appreciated.
    Do I need any special hardware/software on the 6509s and/or on the servers?
    Thanks, in advance.
    John

    A lot depends on the kind of servers you have and the NIC teaming algorithms they support. Usually two or more NICs can be teamed in either a fault tolerant configuration (with one primary NIC) or in transmit load balancing configuration. Both these configurations allow the NICs to be connected to separate L2 switches so in case your servers do support such kind of teaming all you need is to have both NICs in the same VLAN and ensure L2 connectivity between the Catalyst 6509s. I recommend you research this from the server perspective also and like everything else test it before production deployment.

  • Dual boot for lion and SL

    BHow do I set up the dual boot for Lion and snow leopard oon the same internal hard disK
    What does Time machine do  under the dual boot?

    I am also considering putting Lion and SL on separate hard drives.
    Since this is an iMac forum, I assume you will use an external drive as the second drive. Use one that connects via firewire, not via USB.
    Both on a single drive -
    OK. Doing it on a single drive means you will need to partition the internal drive. Although this is straight forward, it is always best to back up the existing drive first, just in case. It's a good idea to have a backup on a different drive, anyway.
    To do that, clone the entire internal drive to the external drive.
    I would suggest, though, that you first partition the external drive. I would suggest four equal-size partitions, provided the drive is big enough - at least 1.0TB, though a 2.0TB drive might be better. You can use OS X's disk utility to do that. Go to the Partition page in Disk Utility, select your external driv in the list on the left, then use the Partition Layout menu to select the number of partitions. Set the format to Mac OS Extended (journaled) GUID for all partitions. Click the Apply button.
    The easy way to do the cloning is to use a cloning utility such as Carbon Copy Cloner. Clone the entire internal drive to one of the partitions on the external drive. Then test boot to it (use Startup Disk in System Preferences to do that) to verify the integrity of the clone. At that time I would suggest you change the background design/picture for the desktop to aomething other than the one used on the original drive - this can be a handy visual reminder of which volume you are booted to.
    Then add a partition to the internal drive. To do that, run Disk Utility. Go to the Partition page, select your internal drive in the list on the left, but do not use the Partition Layout menu this time. Instead, click the existing volume in the graphic display, then click the plus ("+") button underneath that display -
    This will add a partition without erasing the existing information - i.e., the drive will end up with two partitions, one of which will contain your original boot volume.
    Use Startup Disk to re-select your original boot violume, and restart back into it.
    Now, clone the original drive again using CCC, but this time to the second partition of the internal drive, the one you just added.
    Boot to that freshly cloned volume to verify it. Then, while still booted to it, install OS X 10.7 Lion onto it.
    This will result in your original Snow Leopard on the first partition of the internal drive, and Lion on the second partition of the internal drive.
    The external drive will have Snow Leopard on one partition, one partition will be empty (available for a clone of Lion if you wish, or use it for extra storage). The remaining two partitions can be used for Time Machine backups - one for Snow Leopard, one for Lion.
    If you want to have each OS on its own hard drive, then you can use the first clone of Snow Leopard to the external to be upgraded to Lion, or upgrade the orignal on the internal drive to Lion. Be sure to boot to the volume to be used for Lion before installing it so that Lion is installed onto the correct boot volume.
    You can then use the empty partition on the internal drive as the backup (or Time Machine) volume for the OS on the external drive, and vice versa. In this arrangement you may need only partition the external into two volumes.
    If you have a different scheme in mind for partitioning, feel free to do that. I don't use Time Machine, so don't have the need to make space for it. Carbon Copy Cloner can be used to do incremental backups after the initial backup - but I don't use that, either. I'm old school - after the inital cloning for backup, I do manual backups of files on the fly.
    Comments -
    * I would suggest keeping both Snow Leopard and Lion on the internal drive. Reason - even though firewire 800 is fast, it is still noticeably slower than an internal drive. OS's on an external drive will run slower than those on an internal drive.
    • I don't like the concept of Time Machine. Amongst other things, it seems to want more space on the target volume than it actually needs; and it is not a bootable replication. A clone of a bootable volume is bootable.
    • It was just a few days ago that I went through the same process as you are planning. One slight difference - I used a 1.0TB external partitioned into 3 equal volumes, since I had no need to allow space for Time Machine. I have Snow Leopard and Lion each on its own partition of the internal drive; and cloned backups of each on the external drive, plus one GP partition on the external drive.
    • If you have not yet downloaded Lion, be prepared for it to take a while; it's almost 4GB in size. Probably be a good idea to defeat sleep for the duration.
    • Cloning takes a while, but not that long - it took 30 minutes to clone 45GB to the external drive, and 45 minutes to clone it to the second internal partition.

  • BI dual landscape for upgrades

    We recently built a dual / parallel landscape for our BI system to support a support stack upgrade.   I've used the dual landscape concept for many years when doing ERP upgrades, and it has worked very well.   We thought a dual landscape for BI would work also.   However, we did not consider the automatically generated objects in BI.
    What we found is that with 2 development systems, even though BI objects are created with the same logical name, the technical (generated) name will be different between the 2 systems.
    Has anyone successfully implemented a dual-landscape for BI?   If so, how did you syncronize the technical names of objects between development systems without transporting the objects from one development system to the other?   Note: I'm of the opinion that you need to be extremely careful transporting objects between systems with different release/support pack levels and therefore avoid it at all costs . . .  Note 1090842 + several bad experiences.   Therefore, I am very unlikely to consider transporting the objects created in PSD to RLD as a valid solution.
    The following is just an attempt to better describe the issue:
    Production support landscape looks like this:
    PSD  ->  BWP   (these systems have lower release/SP levels than the systems in the release landscape)
    Release landscape looks like this:
    RLD  ->  RLQ    -
    >  Changes will eventually go to BWP after release/SP upgrade in BWP
    Lets say queryA must be developed and transported to BWP prior to the release/SP upgrade in BWP.
    In PSD queryA is created and gets assigned technical name 123456 and is transported to BWP.
    In RLD queryA also gets created, but gets assigned technical name ABCDEF.
    Sometime after the release/SP upgrade, when you are back to a single landscape (RLD -> RLQ -> BWP), a change needs to be made to queryA.   When you transport queryA to BWP, you end up with 2 queryA's, one with technical name 123456 and one with technical name ABCDEF.

    Hey.  I worked for years in a dual landscape environment and at one point, some folks were actually looking at three paths all pointing to production.  The three path idea got squashed fairly quickly, but we had two landscapes for as long as I can remember.  So we essentially had:
    PAD --> PAQ --> Prod
    PBD --> PBQ --> Prod
    This was very helpful to be able to work on major releases and upgrades while still allowing weekly maintenance releases.  I never looked for a document that deeply discussed the concept and pitfalls to watch out for. 
    I will say the system synch-up could often become quite a pain.  This may sound a little crazy but we would transport from landscape A to landscape B but dual code from B to A.  This was because A was for the triannual major releases and upgrades and B was maintenance work.  There were times when we would transport both ways, but we tried to stay away from that (except for queries). 
    The idea was that anything that went to prod as maintnenance, out of B, should be dual coded in parellel in A.  Then once a tri-annual type release went from A, all those new transports and dual coding would be moved to B.  This is how we kept everything in synhc and it worked rather well.  You always had to be careful around the time of the tri-annual go live, but that's to be expected. 
    The only objects we had constant problems with werre queries.  I don't know of anything you can do about that.  We eventually decided to always transport queries between the two landscapes.  We would never manually dual code a query.  We would eventually get the duplicate query issue in production.  The pitfall here is if the same query is getting touched for maintenance and a major release... that was just something we had to workaround.  But we never had any major problems with transporting between the two landscapes.  The times it became a huge headache is when people would forget to dual code, or just plain to a bunch of prototyping in one of these environments when they should have been using a sandbox.
    Not sure if I have answered your question, but hope this helps in some way.  At least you know plenty of us feel your pain!

  • Will the single or dual quad be right?

    Does some have experience to advise me, which model, the single or dual quad "new" Mac Pro will deliver better results for editing images in Photoshop, Lightroom, Capture NX etc. And how much ram is a minimum for same with the new DDR3 type? Also, I do part time video editing (Final Cut Express). Thanks.

    The cpu is smarter than that, and smarter than we are. A lot of things are going on below the surface, and in less than a nono-second is even a stretch in time, and changes. A core doesn't shutdown.
    I suggest you actually dig into the technical papers and articles on Nehalem if you want to understand it more thoroughly. I've read that ARS article three times and still get something new with each read. And there are other articles. I feel like 2008 was just waiting and researching the sucessor to Penryn class (Harpertown).
    The specials are just that, and they get 'hand tuned' and physically booted before they go out the door and usually have above the base spec listed.
    I am not in favor of bleeding edge. Not until there has been one firmware update and one OS patch and time for 3-4 months minimum to flush out problems that surface. Some are still hoping the next OS update fixes issues with Nvidia 8800GT a year after it came out (and yes, drivers take a year of tweaking, Vista has had two years and apps/games and drivers are an ongoing experiment).

  • PC1066 RAMBUS Single or Dual

    Can anyone tell me which is the best and the fastest PC1066 RAMBUS (Single 232pin or Dual 184pin) for the Intel P4 processor.
    Also do MSI make a motherboard that supports Dual P4 processors and PC1066 RAMBUS (232pin or 184pin) and if there is such a moterboard can anyone point me in the right direction.

    RIMM 4200 is the same in preformance terms as dual PC1066.
    Dual P4 doen't exist, you need dual Xeons, preferably with an E7501 or E7505 chipset (dual channel DDR200/266).

  • 2 Switch stack Design for redundancy

    Hi Everyone,
    I need to connect 2 switches in stack which will connect to 2 servers.
    Each server will habe 12 NICs.
    So for Redundancy purposes i can connect 6 ports from server A to Switch 1
                                                                   6 Ports from server A to Switch 2
    Same way i can do
    6 ports from Server B to Switch 1
    6 ports from Server B to switch 2
    IF i go with above design and in case say switch 2 dies then stack will have single switch will it cause any outage?
    Regards
    Mahesh

    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
    IF i go with above design and in case say switch 2 dies then stack will have single switch will it cause any outage?
    It depends.
    One of your two switches will run in the role of switch master, and if that switch fails, there's a brief impact as control plane function are taken over by the second switch.  Whether this control plane switch over causes any impact to the server hosts depends on switch configuration and how (via L2 or L3) the hosts are communicating to other hosts.  For example, if you're running default, regarding master switch MAC persistence, that MAC will change, which I understand, will drop Etherchannel.

  • How do I set-up my TV as a dual monitor for my MacBook Pro?

    Hi
    I have a mid-2009 MacBook Pro running (OS X 10.8.2) Mountain Lion, and recently I rearranged my home workspace so that I can view my TV from my desk while I work.
    I would like to set-up my TV as a dual monitor for my MacBook Pro, so I can stream videos on my TV screen and have the whole screen to do work on my MacBook Pro at the same time.
    Is this possible? I've seen people use TVs as a mirror display (where it works as their main and only monitor), but never as a fully operational dual display.
    As an example, I would like to have my MacBook screen showing a desktop with the whole screen devoted to Photoshop, while I have another desktop running on my TV, which shows an internet browser or other application.
    If this is possible, what cords/hardware would I need, and what are the settings that will enable me to do this?
    - My TV is a modern flatscreen LCD, and it has two HDMI ports in the back.
    - It's about 9-10 feet from my desk.
    If anyone could help, that'd be awesome!
    Thanks.

    Yes, you can do that.
    You'll need a mini displayport to HDMI adaptor:
    http://store.apple.com/us/product/H1824ZM/A/moshi-mini-dp-to-hdmi-adapter-with-a udio-support
    and an HDMI cable long enough to reach your TV
    (look on Amazon or Monoprice for the best price on any of those parts)
    In your display options, you will just choose extended desktop vs mirrored display.

  • How can I use two single-dimensional arrays-one for the titles and array

    I want to Use two single-dimensional arrays-one for the titles and one for the ID
    Could everyone help me how can i write the code for it?
    Flower
    public class Video
    public static void main(String[] args) throws Exception
    int[][] ID =
    { {145,147,148},
    {146,149, 150} };
    String[][] Titles=
    { {"Barney","True Grit","The night before Christmas"},
    {"Lalla", "Jacke Chan", "Metal"} };
    int x, y;
    int r, c;
    System.out.println("List before Sort");
    for(c =0; c< 3; ++c)
    for(r=0; r< 3; ++ r)
    System.out.println("ID:" + ID[c][r]+ "\tTitle: " + Titles[c][r]);
    System.out.println("\nAfter Sort:");
    for(c =0; c< 3; ++c)
    for(r=0; r< 3; ++ r)
    System.out.println("ID:" + ID[c][r]+ "\tTitle: " + Titles[c][r]);

    This is one of the most bizarre questions I have seen here:
    public class Video
    public static void main(String[] args) throws Exception
    int[] ID = {145,147,148, 146,149, 150};
    String[] Titles= {"Barney","True Grit","The night before Christmas", "Lalla", "Jacke Chan", "Metal"};
    System.out.println("List before Sort");
    for(int i = 0; i < Titles.length; i++)
       System.out.println("ID:" + ID[i]+ "\tTitle: " + Titles);
    System.out.println("\nAfter Sort:");
    for(int i = 0; c < Titles.length; i++)
    System.out.println("ID:" + ID[i]+ "\tTitle: " + Titles[i]);
    Generally you don't use prefix (++c) operators in you for loop. Use postfix (c++).
    Prefix means that it will increment the variable before the loop body is executed. Postfix will cause it to increment after.

  • Can't get windows 8.1 with dual boot for fedora (linux) . system is UEFI .

    Can't get  windows 8.1 with  dual boot for fedora (linux) . system is UEFI .

    Hi,
    Any update here?
    We may seek help at Fedora forum as Milos suggested, if convenient we could share the related thread link here for reference.
    In addition, please also check the information in the similar thread:
    Dual Boot Windows 8 and Linux?
    Best regards
    Michael Shao
    TechNet Community Support

  • I'm trying to set up a dual monitor for my new iMac. I'm using a second display I used with my old G5. My iMac recognizes the second monitor but the monitor won't turn on. Power button doesn't do anything on the monitor itself.

    I'm trying to set up a dual monitor for my new iMac. I'm using a second display I used with my old G5. My iMac recognizes the second monitor but the monitor won't turn on. Power button doesn't do anything on the monitor itself.

    If it won't turn on then the monitor is dead (power supply?, video circuits?).  Not worth repairing.
    If you have a new iMac then it has thunderbolt.  So exactly what is that monitor you are trying to connect?  For TB you can connect a mini-displayport monitor.   Probably (not 100% sure with TB) can use dvi or vga too if you have the appropriate adapter.

  • Dual monitors for my macbook pro

    I want to use dual monitors for my macbook pro but I am getting mixed answers as to if that is possible. I want to use just the two monitors, so basically I am thinking I will plug my macbook into the monitors, close my macbook and use a blue tooth mouse and keyboard on the two monitors. I want to know if that is possible. Also, I am wondering about the adapters. I know that one end should be the thunderbolt and then the other end will have whatever the monitor needs (DVI/VGA) but all the adapters I see only have one place to plug a monitor in, and I will be need two.

    It is possible.
    I currently use a thunderbolt to vga adaptor and a usb to vga adaptor to run two independent monitors.
    However, the usb to vga adaptor is not great.  I'm using an Eclipse See2 UV159 but the screen does not come out as good as the thunderbolt to vga adaptor. 
    I'm currently looking for a better adaptor.

  • Does 10.5.6 Support Dual displays for Radeon 9800 Pro Mac Edition cards?

    Does 10.5.6 Support Dual displays for Radeon 9800 Pro Mac Edition cards?
    And is there driver updates for anyone using the all-in-ones HP printers...has anyone been able to use their scanners since the update? and has the dual screen support been added to avoid the blue screen on startup?

    The ATI Radeon 9600 PC/Mac edition 256 MB should work in the MDD G4. It has two DVI ports (one is dual-link and can support a 30" display). You will need adapters for ADC displays.

Maybe you are looking for